.\" DO NOT MODIFY THIS FILE!  It was generated by help2man 1.47.4.
.TH WB_COMMAND "1" "December 2016" "wb_command 1.2.3+git3" "User Commands"
.SH NAME
wb_command \- command-line program for performing a variety of algorithmic tasks using volume, surface, and grayordinate data
.SH SYNOPSIS
.B <class-name>

.SH DESCRIPTION
\fB\-add\-to\-spec\-file\fR
ADD A FILE TO A SPECIFICATION FILE
.IP
wb_command \fB\-add\-to\-spec\-file\fR
.IP
<specfile> \- the specification file to add to
<structure> \- the structure of the data file
<filename> \- the path to the file
.TP
The resulting spec file overwrites the existing spec file.
If the spec
.TP
file doesn't exist, it is created with default metadata.
The structure
.IP
argument must be one of the following:
.IP
CORTEX_LEFT
CORTEX_RIGHT
CEREBELLUM
ACCUMBENS_LEFT
ACCUMBENS_RIGHT
ALL_GREY_MATTER
ALL_WHITE_MATTER
AMYGDALA_LEFT
AMYGDALA_RIGHT
BRAIN_STEM
CAUDATE_LEFT
CAUDATE_RIGHT
CEREBELLAR_WHITE_MATTER_LEFT
CEREBELLAR_WHITE_MATTER_RIGHT
CEREBELLUM_LEFT
CEREBELLUM_RIGHT
CEREBRAL_WHITE_MATTER_LEFT
CEREBRAL_WHITE_MATTER_RIGHT
CORTEX
DIENCEPHALON_VENTRAL_LEFT
DIENCEPHALON_VENTRAL_RIGHT
HIPPOCAMPUS_LEFT
HIPPOCAMPUS_RIGHT
INVALID
OTHER
OTHER_GREY_MATTER
OTHER_WHITE_MATTER
PALLIDUM_LEFT
PALLIDUM_RIGHT
PUTAMEN_LEFT
PUTAMEN_RIGHT
THALAMUS_LEFT
THALAMUS_RIGHT
.PP
\fB\-backend\-average\-dense\-roi\fR
CONNECTOME DB BACKEND COMMAND FOR CIFTI AVERAGE DENSE ROI
.IP
wb_command \fB\-backend\-average\-dense\-roi\fR
.IP
<index\-list> \- comma separated list of cifti indexes to average
<out\-file> \- file to write the average row to
.IP
This command is probably not the one you are looking for, try
\fB\-cifti\-average\-dense\-roi\fR.  It takes the list of cifti files to average
from standard input, and writes its output as little endian, 32\-bit
integer of row size followed by the row as 32\-bit floats.
.PP
\fB\-backend\-average\-roi\-correlation\fR
CONNECTOME DB BACKEND COMMAND FOR CIFTI AVERAGE ROI CORRELATION
.IP
wb_command \fB\-backend\-average\-roi\-correlation\fR
.IP
<index\-list> \- comma separated list of cifti indexes to average and then
.IP
correlate
.IP
<out\-file> \- file to write the average row to
.IP
This command is probably not the one you are looking for, try
\fB\-cifti\-average\-roi\-correlation\fR.  It takes the list of cifti files to
average from standard input, and writes its output as little endian,
32\-bit integer of row size followed by the row as 32\-bit floats.
.PP
\fB\-border\-export\-color\-table\fR
WRITE BORDER NAMES AND COLORS AS TEXT
.IP
wb_command \fB\-border\-export\-color\-table\fR
.IP
<border\-file> \- the input border file
<table\-out> \- output \- the output text file
.IP
[\-class\-colors] \- use class colors instead of the name colors
.IP
Takes the names and colors of each border, and writes it to the same
format as \fB\-metric\-label\-import\fR expects.  By default, the borders are
colored by border name, specify \fB\-class\-colors\fR to color them by class
instead.  The key values start at 1 and follow the order of the borders
in the file.
.PP
\fB\-border\-file\-export\-to\-caret5\fR
EXPORT BORDER FILE TO CARET5 FILE FORMAT
.IP
wb_command \fB\-border\-file\-export\-to\-caret5\fR
.IP
<border\-file> \- workbench border file
<output\-file\-prefix> \- prefix for name of output caret5
.IP
border/borderproj/bordercolor files
.IP
[\-surface] \- repeatable \- specify an input surface
.IP
<surface\-in> \- a surface file for unprojection of borders
.IP
A Workbench border file may contain borders for multiple structures and
borders that are both projected and unprojected.  It also contains a
color table for the borders.
.IP
Caret5 has both border (unprojected) and border projection (projected)
files.  In addition, each Caret5 border or border projection file
typically contains data for a single structure. Caret5 also uses a border
color file that associates colors with the names of the borders.
.IP
This command will try to output both Caret5 border and border projection
files.  Each output border/border projection file will contains data for
one structure so there may be many files created.  The structure name is
included in the name of each border or border projection file that is
created.
.IP
One Caret5 border color file will also be produced by this command.
.IP
Providing surface(s) as input parameters is optional, but recommended.
Surfaces may be needed to create both projected and/or unprojected
coordinates of borders.  If there is a failure to produce an output
border or border projection due to a missing surface with the matching
structure, an error message will be displayed and some output files will
not be created.
.IP
When writing new files, this command will overwrite a file with the same
name.
.PP
\fB\-border\-length\fR
REPORT LENGTH OF BORDERS
.IP
wb_command \fB\-border\-length\fR
.IP
<border> \- the input border file
<surface> \- the surface to measure the borders on
.IP
[\-corrected\-areas] \- vertex areas to use instead of computing them from
.IP
the surface
<area\-metric> \- the corrected vertex areas, as a metric
.IP
[\-separate\-pieces] \- report lengths for multi\-part borders as separate
.IP
numbers
.IP
[\-hide\-border\-name] \- don't print border name before each output
.TP
For each border, print its length along the surface, in mm.
If a border
.IP
has multiple parts, their lengths are summed before printing, unless
\fB\-separate\-pieces\fR is specified.
.IP
The \fB\-corrected\-areas\fR option is intended for when the length is not
meaningfully measurable on individual surfaces, it is only an approximate
correction for the reduction in structure of a group average surface.
.PP
\fB\-border\-merge\fR
MERGE BORDER FILES INTO A NEW FILE
.IP
wb_command \fB\-border\-merge\fR
.IP
<border\-file\-out> \- output \- the output border file
.IP
[\-border] \- repeatable \- specify an input border file
.IP
<border\-file\-in> \- a border file to use borders from
.IP
[\-select] \- repeatable \- select a single border to use
.IP
<border> \- the border number or name
.IP
[\-up\-to] \- use an inclusive range of borders
.IP
<last\-border> \- the number or name of the last column to include
.IP
[\-reverse] \- use the range in reverse order
.IP
Takes one or more border files and makes a new border file from the
borders in them.
.IP
Example: wb_command \fB\-border\-merge\fR out.border \fB\-border\fR first.border \fB\-select\fR
1 \fB\-border\fR second.border
.IP
This example would take the first border from first.border, followed by
all borders from second.border, and write these to out.border.
.PP
\fB\-border\-resample\fR
RESAMPLE A BORDER FILE TO A DIFFERENT MESH
.IP
wb_command \fB\-border\-resample\fR
.IP
<border\-in> \- the border file to resample
<current\-sphere> \- a sphere surface with the mesh that the metric is
.IP
currently on
.IP
<new\-sphere> \- a sphere surface that is in register with <current\-sphere>
.IP
and has the desired output mesh
.IP
<border\-out> \- output \- the output border file
.IP
Resamples a border file, given two spherical surfaces that are in
register.  Only borders that have the same structure as current\-sphere
will be resampled.
.PP
\fB\-border\-to\-rois\fR
MAKE METRIC ROIS FROM BORDERS
.IP
wb_command \fB\-border\-to\-rois\fR
.IP
<surface> \- the surface the borders are drawn on
<border\-file> \- the border file
<metric\-out> \- output \- the output metric file
.IP
[\-border] \- create ROI for only one border
.IP
<name> \- the name of the border
.IP
[\-inverse] \- use inverse selection (outside border)
.IP
By default, draws ROIs inside all borders in the border file, as separate
metric columns.
.PP
\fB\-border\-to\-vertices\fR
DRAW BORDERS AS VERTICES IN A METRIC FILE
.IP
wb_command \fB\-border\-to\-vertices\fR
.IP
<surface> \- the surface the borders are drawn on
<border\-file> \- the border file
<metric\-out> \- output \- the output metric file
.IP
[\-border] \- create ROI for only one border
.IP
<name> \- the name of the border
.IP
Outputs a metric with 1s on vertices that follow a border, and 0s
elsewhere.  By default, a separate metric column is created for each
border.
.PP
\fB\-cifti\-all\-labels\-to\-rois\fR
MAKE ROIS FROM ALL LABELS IN A CIFTI LABEL MAP
.IP
wb_command \fB\-cifti\-all\-labels\-to\-rois\fR
.IP
<label\-in> \- the input cifti label file
<map> \- the number or name of the label map to use
<cifti\-out> \- output \- the output cifti file
.IP
The output cifti file is a dscalar file with a column (map) for each
label in the specified input map, other than the ??? label, each of which
contains a binary ROI of all brainordinates that are set to the
corresponding label.
.IP
Most of the time, specifying '1' for the <map> argument will do what is
desired.
.PP
\fB\-cifti\-average\fR
AVERAGE CIFTI FILES
.IP
wb_command \fB\-cifti\-average\fR
.IP
<cifti\-out> \- output \- output cifti file
.IP
[\-exclude\-outliers] \- exclude outliers by standard deviation of each
.IP
element across files
<sigma\-below> \- number of standard deviations below the mean to
.IP
include
.IP
<sigma\-above> \- number of standard deviations above the mean to
.IP
include
.IP
[\-cifti] \- repeatable \- specify an input file
.IP
<cifti\-in> \- the input cifti file
.IP
[\-weight] \- give a weight for this file
.IP
<weight> \- the weight to use
.TP
Averages cifti files together.
Files without \fB\-weight\fR specified are given
.TP
a weight of 1.
If \fB\-exclude\-outliers\fR is specified, at each element, the
.IP
data across all files is taken as a set, its unweighted mean and sample
standard deviation are found, and values outside the specified number of
standard deviations are excluded from the (potentially weighted) average
at that element.
.PP
\fB\-cifti\-average\-dense\-roi\fR
AVERAGE CIFTI ROWS ACROSS SUBJECTS BY ROI
.IP
wb_command \fB\-cifti\-average\-dense\-roi\fR
.IP
<cifti\-out> \- output \- output cifti dscalar file
.IP
[\-cifti\-roi] \- cifti file containing combined weights
.IP
<roi\-cifti> \- the roi cifti file
.IP
[\-in\-memory] \- cache the roi in memory so that it isn't re\-read for
.IP
each input cifti
.IP
[\-left\-roi] \- weights to use for left hempsphere
.IP
<roi\-metric> \- the left roi as a metric file
.IP
[\-right\-roi] \- weights to use for right hempsphere
.IP
<roi\-metric> \- the right roi as a metric file
.IP
[\-cerebellum\-roi] \- weights to use for cerebellum surface
.IP
<roi\-metric> \- the cerebellum roi as a metric file
.IP
[\-vol\-roi] \- voxel weights to use
.IP
<roi\-vol> \- the roi volume file
.IP
[\-left\-area\-surf] \- specify the left surface for vertex area correction
.IP
<left\-surf> \- the left surface file
.IP
[\-right\-area\-surf] \- specify the right surface for vertex area correction
.IP
<right\-surf> \- the right surface file
.IP
[\-cerebellum\-area\-surf] \- specify the cerebellum surface for vertex area
.IP
correction
<cerebellum\-surf> \- the cerebellum surface file
.IP
[\-cifti] \- repeatable \- specify an input cifti file
.IP
<cifti\-in> \- a cifti file to average across
.TP
Averages rows for each map of the ROI(s), across all files.
ROI maps are
.TP
treated as weighting functions, including negative values.
For
.IP
efficiency, ensure that everything that is not intended to be used is
zero in the ROI map.  If \fB\-cifti\-roi\fR is specified, \fB\-left\-roi\fR, \fB\-right\-roi\fR,
\fB\-cerebellum\-roi\fR, and \fB\-vol\-roi\fR must not be specified.  If multiple
non\-cifti ROI files are specified, they must have the same number of
columns.
.PP
\fB\-cifti\-average\-roi\-correlation\fR
CORRELATE ROI AVERAGE WITH ALL ROWS THEN AVERAGE ACROSS SUBJECTS
.IP
wb_command \fB\-cifti\-average\-roi\-correlation\fR
.IP
<cifti\-out> \- output \- output cifti file
.IP
[\-cifti\-roi] \- cifti file containing combined weights
.IP
<roi\-cifti> \- the roi cifti file
.IP
[\-in\-memory] \- cache the roi in memory so that it isn't re\-read for
.IP
each input cifti
.IP
[\-left\-roi] \- weights to use for left hempsphere
.IP
<roi\-metric> \- the left roi as a metric file
.IP
[\-right\-roi] \- weights to use for right hempsphere
.IP
<roi\-metric> \- the right roi as a metric file
.IP
[\-cerebellum\-roi] \- weights to use for cerebellum surface
.IP
<roi\-metric> \- the cerebellum roi as a metric file
.IP
[\-vol\-roi] \- voxel weights to use
.IP
<roi\-vol> \- the roi volume file
.IP
[\-left\-area\-surf] \- specify the left surface for vertex area correction
.IP
<left\-surf> \- the left surface file
.IP
[\-right\-area\-surf] \- specify the right surface for vertex area correction
.IP
<right\-surf> \- the right surface file
.IP
[\-cerebellum\-area\-surf] \- specify the cerebellum surface for vertex area
.IP
correction
<cerebellum\-surf> \- the cerebellum surface file
.IP
[\-cifti] \- repeatable \- specify an input cifti file
.IP
<cifti\-in> \- a cifti file to average across
.IP
Averages rows for each map of the ROI(s), takes the correlation of each
ROI average to the rest of the rows in the same file, then averages the
results across all files.  ROIs are always treated as weighting
functions, including negative values.  For efficiency, ensure that
everything that is not intended to be used is zero in the ROI map.  If
\fB\-cifti\-roi\fR is specified, \fB\-left\-roi\fR, \fB\-right\-roi\fR, \fB\-cerebellum\-roi\fR, and
\fB\-vol\-roi\fR must not be specified.  If multiple non\-cifti ROI files are
specified, they must have the same number of columns.
.PP
\fB\-cifti\-change\-mapping\fR
CONVERT TO SCALAR, COPY MAPPING, ETC
.IP
wb_command \fB\-cifti\-change\-mapping\fR
.IP
<data\-cifti> \- the cifti file to use the data from
<direction> \- which direction on <data\-cifti> to replace the mapping
<cifti\-out> \- output \- the output cifti file
.IP
[\-series] \- set the mapping to series
.IP
<step> \- increment between series points
<start> \- start value of the series
.IP
[\-unit] \- select unit for series (default SECOND)
.IP
<unit> \- unit identifier
.IP
[\-scalar] \- set the mapping to scalar
.IP
[\-name\-file] \- specify names for the maps
.IP
<file> \- text file containing map names, one per line
.IP
[\-from\-cifti] \- copy mapping from another cifti file
.IP
<template\-cifti> \- a cifti file containing the desired mapping
<direction> \- which direction to copy the mapping from
.TP
Take an existing cifti file and change one of the mappings.
Exactly one
.TP
of \fB\-series\fR, \fB\-scalar\fR, or \fB\-from\-cifti\fR must be specified.
The direction can
.IP
be either an integer starting from 1, or the strings 'ROW' or 'COLUMN'.
.IP
The argument to \fB\-unit\fR must be one of the following:
.IP
SECOND
HERTZ
METER
RADIAN
.PP
\fB\-cifti\-convert\fR
DUMP CIFTI MATRIX INTO OTHER FORMATS
.IP
wb_command \fB\-cifti\-convert\fR
.IP
[\-to\-gifti\-ext] \- convert to GIFTI external binary
.IP
<cifti\-in> \- the input cifti file
<gifti\-out> \- output \- the output gifti file
.IP
[\-from\-gifti\-ext] \- convert a GIFTI made with this command back into a
.IP
CIFTI
<gifti\-in> \- the input gifti file
<cifti\-out> \- output \- the output cifti file
.IP
[\-reset\-timepoints] \- reset the mapping along rows to timepoints,
.IP
taking length from the gifti file
<timestep> \- the desired time between frames
<timestart> \- the desired time offset of the initial frame
.IP
[\-unit] \- use a unit other than time
.IP
<unit> \- unit identifier (default SECOND)
.IP
[\-reset\-scalars] \- reset mapping along rows to scalars, taking length
.IP
from the gifti file
.IP
[\-replace\-binary] \- replace data with a binary file
.IP
<binary\-in> \- the binary file that contains replacement data
.IP
[\-flip\-endian] \- byteswap the binary file
.IP
[\-transpose] \- transpose the binary file
.IP
[\-to\-nifti] \- convert to NIFTI1
.IP
<cifti\-in> \- the input cifti file
<nifti\-out> \- output \- the output nifti file
.IP
[\-from\-nifti] \- convert a NIFTI (1 or 2) file made with this command back
.IP
into CIFTI
<nifti\-in> \- the input nifti file
<cifti\-template> \- a cifti file with the dimension(s) and mapping(s)
.IP
that should be used
.IP
<cifti\-out> \- output \- the output cifti file
.IP
[\-reset\-timepoints] \- reset the mapping along rows to timepoints,
.IP
taking length from the nifti file
<timestep> \- the desired time between frames
<timestart> \- the desired time offset of the initial frame
.IP
[\-unit] \- use a unit other than time
.IP
<unit> \- unit identifier (default SECOND)
.IP
[\-reset\-scalars] \- reset mapping along rows to scalars, taking length
.IP
from the nifti file
.IP
[\-to\-text] \- convert to a plain text file
.IP
<cifti\-in> \- the input cifti file
<text\-out> \- output \- the output text file
.IP
[\-col\-delim] \- choose string to put between elements in a row
.IP
<delim\-string> \- the string to use (default is a tab character)
.IP
[\-from\-text] \- convert from plain text to cifti
.IP
<text\-in> \- the input text file
<cifti\-template> \- a cifti file with the dimension(s) and mapping(s)
.IP
that should be used
.IP
<cifti\-out> \- output \- the output cifti file
.IP
[\-col\-delim] \- specify string that is between elements in a row
.IP
<delim\-string> \- the string to use (default is any whitespace)
.IP
[\-reset\-timepoints] \- reset the mapping along rows to timepoints,
.IP
taking length from the text file
<timestep> \- the desired time between frames
<timestart> \- the desired time offset of the initial frame
.IP
[\-unit] \- use a unit other than time
.IP
<unit> \- unit identifier (default SECOND)
.IP
[\-reset\-scalars] \- reset mapping along rows to scalars, taking length
.IP
from the text file
.IP
This command is used to convert a full CIFTI matrix to/from formats that
can be used by programs that don't understand CIFTI.  You must specify
exactly one of \fB\-to\-gifti\-ext\fR, \fB\-from\-gifti\-ext\fR, \fB\-to\-nifti\fR, \fB\-from\-nifti\fR,
\fB\-to\-text\fR, or \fB\-from\-text\fR.
.IP
If you want to write an existing CIFTI file with a different CIFTI
version, see \fB\-file\-convert\fR, and its \fB\-cifti\-version\-convert\fR option.
.IP
If you want part of the CIFTI file as a metric, label, or volume file,
see \fB\-cifti\-separate\fR.  If you want to create a CIFTI file from metric
and/or volume files, see the \fB\-cifti\-create\-\fR* commands.
.IP
If you want to import a matrix that is restricted to an ROI, first create
a template CIFTI file matching that ROI using a \fB\-cifti\-create\-\fR* command.
After importing to CIFTI, you can then expand the file into a standard
brainordinates space with \fB\-cifti\-create\-dense\-from\-template\fR.  If you want
to export only part of a CIFTI file, first create an roi\-restricted CIFTI
file with \fB\-cifti\-restrict\-dense\-mapping\fR.
.IP
The \fB\-transpose\fR option to \fB\-from\-gifti\-ext\fR is needed if the replacement
binary file is in column\-major order.
.IP
The \fB\-unit\fR options accept these values:
.IP
SECOND
HERTZ
METER
RADIAN
.PP
\fB\-cifti\-correlation\fR
GENERATE CORRELATION OF ROWS IN A CIFTI FILE
.IP
wb_command \fB\-cifti\-correlation\fR
.IP
<cifti> \- input cifti file
<cifti\-out> \- output \- output cifti file
.IP
[\-roi\-override] \- perform correlation from a subset of rows to all rows
.IP
[\-left\-roi] \- use an roi for left hempsphere
.IP
<roi\-metric> \- the left roi as a metric file
.IP
[\-right\-roi] \- use an roi for right hempsphere
.IP
<roi\-metric> \- the right roi as a metric file
.IP
[\-cerebellum\-roi] \- use an roi for cerebellum
.IP
<roi\-metric> \- the cerebellum roi as a metric file
.IP
[\-vol\-roi] \- use an roi for volume
.IP
<roi\-vol> \- the volume roi file
.IP
[\-cifti\-roi] \- use a cifti file for combined rois
.IP
<roi\-cifti> \- the cifti roi file
.IP
[\-weights] \- specify column weights
.IP
<weight\-file> \- text file containing one weight per column
.IP
[\-fisher\-z] \- apply fisher small z transform (ie, artanh) to correlation
.IP
[\-no\-demean] \- instead of correlation, do dot product of rows, then
.IP
normalize by diagonal
.IP
[\-covariance] \- compute covariance instead of correlation
.IP
[\-mem\-limit] \- restrict memory usage
.IP
<limit\-GB> \- memory limit in gigabytes
.IP
For each row (or each row inside an roi if \fB\-roi\-override\fR is specified),
correlate to all other rows.  The \fB\-cifti\-roi\fR suboption to \fB\-roi\-override\fR
may not be specified with any other \-*\-roi suboption, but you may specify
the other \-*\-roi suboptions together.
.IP
When using the \fB\-fisher\-z\fR option, the output is NOT a Z\-score, it is
artanh(r), to do further math on this output, consider using \fB\-cifti\-math\fR.
.IP
Restricting the memory usage will make it calculate the output in chunks,
and if the input file size is more than 70% of the memory limit, it will
also read through the input file as rows are required, resulting in
several passes through the input file (once per chunk).  Memory limit
does not need to be an integer, you may also specify 0 to calculate a
single output row at a time (this may be very slow).
.PP
\fB\-cifti\-correlation\-gradient\fR
CORRELATE CIFTI ROWS AND TAKE GRADIENT
.IP
wb_command \fB\-cifti\-correlation\-gradient\fR
.IP
<cifti> \- the input cifti
<cifti\-out> \- output \- the output cifti
.IP
[\-left\-surface] \- specify the left surface to use
.IP
<surface> \- the left surface file
.IP
[\-left\-corrected\-areas] \- vertex areas to use instead of computing
.IP
them from the left surface
<area\-metric> \- the corrected vertex areas, as a metric
.IP
[\-right\-surface] \- specify the right surface to use
.IP
<surface> \- the right surface file
.IP
[\-right\-corrected\-areas] \- vertex areas to use instead of computing
.IP
them from the right surface
<area\-metric> \- the corrected vertex areas, as a metric
.IP
[\-cerebellum\-surface] \- specify the cerebellum surface to use
.IP
<surface> \- the cerebellum surface file
.IP
[\-cerebellum\-corrected\-areas] \- vertex areas to use instead of
.IP
computing them from the cerebellum surface
<area\-metric> \- the corrected vertex areas, as a metric
.IP
[\-surface\-presmooth] \- smooth on the surface before computing the
.IP
gradient
<surface\-kernel> \- the sigma for the gaussian surface smoothing
.IP
kernel, in mm
.IP
[\-volume\-presmooth] \- smooth the volume before computing the gradient
.IP
<volume\-kernel> \- the sigma for the gaussian volume smoothing kernel,
.IP
in mm
.IP
[\-undo\-fisher\-z] \- apply the inverse fisher small z transform to the
.IP
input
.IP
[\-fisher\-z] \- apply the fisher small z transform to the correlations
.IP
before taking the gradient
.IP
[\-surface\-exclude] \- exclude vertices near each seed vertex from
.IP
computation
<distance> \- geodesic distance from seed vertex for the exclusion
.IP
zone, in mm
.IP
[\-volume\-exclude] \- exclude voxels near each seed voxel from computation
.IP
<distance> \- distance from seed voxel for the exclusion zone, in mm
.IP
[\-covariance] \- compute covariance instead of correlation
.IP
[\-mem\-limit] \- restrict memory usage
.IP
<limit\-GB> \- memory limit in gigabytes
.IP
For each structure, compute the correlation of the rows in the structure,
and take the gradients of the resulting rows, then average them.  Memory
limit does not need to be an integer, you may also specify 0 to use as
little memory as possible (this may be very slow).
.PP
\fB\-cifti\-create\-dense\-from\-template\fR
CREATE CIFTI WITH MATCHING DENSE MAP
.IP
wb_command \fB\-cifti\-create\-dense\-from\-template\fR
.IP
<template\-cifti> \- file to match brainordinates of
<cifti\-out> \- output \- the output cifti file
.IP
[\-series] \- make a dtseries file instead of a dscalar
.IP
<step> \- increment between series points
<start> \- start value of the series
.IP
[\-unit] \- select unit for series (default SECOND)
.IP
<unit> \- unit identifier
.IP
[\-volume\-all] \- specify an input volume file for all voxel data
.IP
<volume\-in> \- the input volume file
.IP
[\-from\-cropped] \- the input is cropped to the size of the voxel data
.IP
in the template file
.IP
[\-cifti] \- repeatable \- use input data from a cifti file
.IP
<cifti\-in> \- cifti file containing input data
.IP
[\-metric] \- repeatable \- use input data from a metric file
.IP
<structure> \- which structure to put the metric file into
<metric\-in> \- input metric file
.IP
[\-label] \- repeatable \- use input data from surface label files
.IP
<structure> \- which structure to put the label file into
<label\-in> \- input label file
.IP
[\-volume] \- repeatable \- use a volume file for a single volume
.IP
structure's data
<structure> \- which structure to put the volume file into
<volume\-in> \- the input volume file
.IP
[\-from\-cropped] \- the input is cropped to the size of the volume
.IP
structure
.IP
This command helps you make a new dscalar, dtseries, or dlabel cifti file
that matches the brainordinate space used in another cifti file.  The
template file must have the desired brainordinate space in the mapping
along the column direction (for dtseries, dscalar, dlabel, and symmetric
dconn this is always the case).  All input cifti files must have a brain
models mapping along column and use the same volume space and/or surface
vertex count as the template for structures that they contain.  If any
input files contain label data, then input files with non\-label data are
not allowed, and the \fB\-series\fR option may not be used.
.IP
Any structure that isn't covered by an input is filled with zeros or the
unlabeled key.
.IP
The <structure> argument of \fB\-metric\fR, \fB\-label\fR or \fB\-volume\fR must be one of the
following:
.IP
CORTEX_LEFT
CORTEX_RIGHT
CEREBELLUM
ACCUMBENS_LEFT
ACCUMBENS_RIGHT
ALL_GREY_MATTER
ALL_WHITE_MATTER
AMYGDALA_LEFT
AMYGDALA_RIGHT
BRAIN_STEM
CAUDATE_LEFT
CAUDATE_RIGHT
CEREBELLAR_WHITE_MATTER_LEFT
CEREBELLAR_WHITE_MATTER_RIGHT
CEREBELLUM_LEFT
CEREBELLUM_RIGHT
CEREBRAL_WHITE_MATTER_LEFT
CEREBRAL_WHITE_MATTER_RIGHT
CORTEX
DIENCEPHALON_VENTRAL_LEFT
DIENCEPHALON_VENTRAL_RIGHT
HIPPOCAMPUS_LEFT
HIPPOCAMPUS_RIGHT
INVALID
OTHER
OTHER_GREY_MATTER
OTHER_WHITE_MATTER
PALLIDUM_LEFT
PALLIDUM_RIGHT
PUTAMEN_LEFT
PUTAMEN_RIGHT
THALAMUS_LEFT
THALAMUS_RIGHT
.IP
The argument to \fB\-unit\fR must be one of the following:
.IP
SECOND
HERTZ
METER
RADIAN
.PP
\fB\-cifti\-create\-dense\-scalar\fR
CREATE A CIFTI DENSE SCALAR FILE
.IP
wb_command \fB\-cifti\-create\-dense\-scalar\fR
.IP
<cifti\-out> \- output \- the output cifti file
.IP
[\-volume] \- volume component
.IP
<volume\-data> \- volume file containing all voxel data for all volume
.IP
structures
.IP
<label\-volume> \- label volume file containing labels for cifti
.IP
structures
.IP
[\-left\-metric] \- metric for left surface
.IP
<metric> \- the metric file
.IP
[\-roi\-left] \- roi of vertices to use from left surface
.IP
<roi\-metric> \- the ROI as a metric file
.IP
[\-right\-metric] \- metric for left surface
.IP
<metric> \- the metric file
.IP
[\-roi\-right] \- roi of vertices to use from right surface
.IP
<roi\-metric> \- the ROI as a metric file
.IP
[\-cerebellum\-metric] \- metric for the cerebellum
.IP
<metric> \- the metric file
.IP
[\-roi\-cerebellum] \- roi of vertices to use from right surface
.IP
<roi\-metric> \- the ROI as a metric file
.IP
[\-name\-file] \- use a text file to set all map names
.IP
<file> \- text file containing map names, one per line
.TP
All input files must have the same number of columns/subvolumes.
Only
.TP
the specified components will be in the output cifti.
Map names will be
.TP
taken from one of the input files.
At least one component must be
.TP
specified.
The label volume should have some of the label names from
.IP
this list, all other label names will be ignored:
.IP
CORTEX_LEFT
CORTEX_RIGHT
CEREBELLUM
ACCUMBENS_LEFT
ACCUMBENS_RIGHT
ALL_GREY_MATTER
ALL_WHITE_MATTER
AMYGDALA_LEFT
AMYGDALA_RIGHT
BRAIN_STEM
CAUDATE_LEFT
CAUDATE_RIGHT
CEREBELLAR_WHITE_MATTER_LEFT
CEREBELLAR_WHITE_MATTER_RIGHT
CEREBELLUM_LEFT
CEREBELLUM_RIGHT
CEREBRAL_WHITE_MATTER_LEFT
CEREBRAL_WHITE_MATTER_RIGHT
CORTEX
DIENCEPHALON_VENTRAL_LEFT
DIENCEPHALON_VENTRAL_RIGHT
HIPPOCAMPUS_LEFT
HIPPOCAMPUS_RIGHT
INVALID
OTHER
OTHER_GREY_MATTER
OTHER_WHITE_MATTER
PALLIDUM_LEFT
PALLIDUM_RIGHT
PUTAMEN_LEFT
PUTAMEN_RIGHT
THALAMUS_LEFT
THALAMUS_RIGHT
.PP
\fB\-cifti\-create\-dense\-timeseries\fR
CREATE A CIFTI DENSE TIMESERIES
.IP
wb_command \fB\-cifti\-create\-dense\-timeseries\fR
.IP
<cifti\-out> \- output \- the output cifti file
.IP
[\-volume] \- volume component
.IP
<volume\-data> \- volume file containing all voxel data for all volume
.IP
structures
.IP
<label\-volume> \- label volume file containing labels for cifti
.IP
structures
.IP
[\-left\-metric] \- metric for left surface
.IP
<metric> \- the metric file
.IP
[\-roi\-left] \- roi of vertices to use from left surface
.IP
<roi\-metric> \- the ROI as a metric file
.IP
[\-right\-metric] \- metric for left surface
.IP
<metric> \- the metric file
.IP
[\-roi\-right] \- roi of vertices to use from right surface
.IP
<roi\-metric> \- the ROI as a metric file
.IP
[\-cerebellum\-metric] \- metric for the cerebellum
.IP
<metric> \- the metric file
.IP
[\-roi\-cerebellum] \- roi of vertices to use from right surface
.IP
<roi\-metric> \- the ROI as a metric file
.IP
[\-timestep] \- set the timestep
.IP
<interval> \- the timestep, in seconds (default 1.0)
.IP
[\-timestart] \- set the start time
.IP
<start> \- the time at the first frame, in seconds (default 0.0)
.IP
[\-unit] \- use a unit other than time
.IP
<unit> \- unit identifier (default SECOND)
.TP
All input files must have the same number of columns/subvolumes.
Only
.TP
the specified components will be in the output cifti.
At least one
.TP
component must be specified.
The label volume should have some of the
.IP
label names from this list, all other label names will be ignored:
.IP
CORTEX_LEFT
CORTEX_RIGHT
CEREBELLUM
ACCUMBENS_LEFT
ACCUMBENS_RIGHT
ALL_GREY_MATTER
ALL_WHITE_MATTER
AMYGDALA_LEFT
AMYGDALA_RIGHT
BRAIN_STEM
CAUDATE_LEFT
CAUDATE_RIGHT
CEREBELLAR_WHITE_MATTER_LEFT
CEREBELLAR_WHITE_MATTER_RIGHT
CEREBELLUM_LEFT
CEREBELLUM_RIGHT
CEREBRAL_WHITE_MATTER_LEFT
CEREBRAL_WHITE_MATTER_RIGHT
CORTEX
DIENCEPHALON_VENTRAL_LEFT
DIENCEPHALON_VENTRAL_RIGHT
HIPPOCAMPUS_LEFT
HIPPOCAMPUS_RIGHT
INVALID
OTHER
OTHER_GREY_MATTER
OTHER_WHITE_MATTER
PALLIDUM_LEFT
PALLIDUM_RIGHT
PUTAMEN_LEFT
PUTAMEN_RIGHT
THALAMUS_LEFT
THALAMUS_RIGHT
.IP
The \fB\-unit\fR option accepts these values:
.IP
SECOND
HERTZ
METER
RADIAN
.PP
\fB\-cifti\-create\-label\fR
CREATE A CIFTI LABEL FILE
.IP
wb_command \fB\-cifti\-create\-label\fR
.IP
<cifti\-out> \- output \- the output cifti file
.IP
[\-volume] \- volume component
.IP
<label\-volume> \- volume file containing the label data
<parcel\-volume> \- label volume file with cifti structure names to
.IP
define the volume parcels
.IP
[\-left\-label] \- label file for left surface
.IP
<label> \- the label file
.IP
[\-roi\-left] \- roi of vertices to use from left surface
.IP
<roi\-metric> \- the ROI as a metric file
.IP
[\-right\-label] \- label for left surface
.IP
<label> \- the label file
.IP
[\-roi\-right] \- roi of vertices to use from right surface
.IP
<roi\-metric> \- the ROI as a metric file
.IP
[\-cerebellum\-label] \- label for the cerebellum
.IP
<label> \- the label file
.IP
[\-roi\-cerebellum] \- roi of vertices to use from right surface
.IP
<roi\-metric> \- the ROI as a metric file
.TP
All input files must have the same number of columns/subvolumes.
Only
.TP
the specified components will be in the output cifti.
At least one
.IP
component must be specified.
.IP
The \fB\-volume\fR option of \fB\-cifti\-create\-label\fR requires two volume arguments,
the label\-volume argument contains all labels you want to display (e.g.
nuclei of the thalamus), whereas the parcel\-volume argument includes all
CIFTI structures you want to include data within (e.g. THALAMUS_LEFT,
THALAMUS_RIGHT).  If you just want the labels in voxels to be the
structure names, you may use the same file for both arguments.  The
parcel\-volume must use some of the label names from this list, all other
label names in the parcel\-volume will be ignored:
.IP
CORTEX_LEFT
CORTEX_RIGHT
CEREBELLUM
ACCUMBENS_LEFT
ACCUMBENS_RIGHT
ALL_GREY_MATTER
ALL_WHITE_MATTER
AMYGDALA_LEFT
AMYGDALA_RIGHT
BRAIN_STEM
CAUDATE_LEFT
CAUDATE_RIGHT
CEREBELLAR_WHITE_MATTER_LEFT
CEREBELLAR_WHITE_MATTER_RIGHT
CEREBELLUM_LEFT
CEREBELLUM_RIGHT
CEREBRAL_WHITE_MATTER_LEFT
CEREBRAL_WHITE_MATTER_RIGHT
CORTEX
DIENCEPHALON_VENTRAL_LEFT
DIENCEPHALON_VENTRAL_RIGHT
HIPPOCAMPUS_LEFT
HIPPOCAMPUS_RIGHT
INVALID
OTHER
OTHER_GREY_MATTER
OTHER_WHITE_MATTER
PALLIDUM_LEFT
PALLIDUM_RIGHT
PUTAMEN_LEFT
PUTAMEN_RIGHT
THALAMUS_LEFT
THALAMUS_RIGHT
.PP
\fB\-cifti\-create\-parcellated\-from\-template\fR
MATCH PARCELS TO TEMPLATE BY NAME
.IP
wb_command \fB\-cifti\-create\-parcellated\-from\-template\fR
.IP
<cifti\-template> \- a cifti file with the template parcel mapping along
.IP
column
.IP
<modify\-direction> \- which dimension of the output file should match the
.IP
template (integer, 'ROW', or 'COLUMN')
.IP
<cifti\-out> \- output \- the output cifti file
.IP
[\-fill\-value] \- specify value to be used in parcels that don't match
.IP
<value> \- value to use (default 0)
.IP
[\-cifti] \- repeatable \- specify an input cifti file
.IP
<cifti\-in> \- the input parcellated cifti file
.IP
For each parcel name in the template mapping, find that name in an input
cifti file and use its data in the output file.  All input cifti files
must have a parcels mapping along <modify\-direction> and matching
mappings along other dimensions.  The direction can be either an integer
starting from 1, or the strings 'ROW' or 'COLUMN'.
.PP
\fB\-cifti\-create\-scalar\-series\fR
IMPORT SERIES DATA INTO CIFTI
.IP
wb_command \fB\-cifti\-create\-scalar\-series\fR
.IP
<input> \- input file
<cifti\-out> \- output \- output cifti file
.IP
[\-transpose] \- use if the rows of the text file are along the scalar
.IP
dimension
.IP
[\-name\-file] \- use a text file to set names on scalar dimension
.IP
<file> \- text file containing names, one per line
.IP
[\-series] \- set the units and values of the series
.IP
<unit> \- the unit to use
<start> \- the value at the first series point
<step> \- the interval between series points
.IP
Convert a text file containing series of equal length into a cifti file.
The text file should have lines made up of numbers separated by
whitespace, with no extra newlines between lines.
.IP
The <unit> argument must be one of the following:
.IP
SECOND
HERTZ
METER
RADIAN
.PP
\fB\-cifti\-cross\-correlation\fR
CORRELATE A CIFTI FILE WITH ANOTHER CIFTI FILE
.IP
wb_command \fB\-cifti\-cross\-correlation\fR
.IP
<cifti\-a> \- first input cifti file
<cifti\-b> \- second input cifti file
<cifti\-out> \- output \- output cifti file
.IP
[\-weights] \- specify column weights
.IP
<weight\-file> \- text file containing one weight per column
.IP
[\-fisher\-z] \- apply fisher small z transform (ie, artanh) to correlation
.IP
[\-mem\-limit] \- restrict memory usage
.IP
<limit\-GB> \- memory limit in gigabytes
.TP
Correlates every rown in <cifti\-a> with every row in <cifti\-b>.
The
.IP
mapping along columns in <cifti\-b> becomes the mapping along rows in the
output.
.IP
When using the \fB\-fisher\-z\fR option, the output is NOT a Z\-score, it is
artanh(r), to do further math on this output, consider using \fB\-cifti\-math\fR.
.IP
Restricting the memory usage will make it calculate the output in chunks,
by reading through <cifti\-b> multiple times.
.PP
\fB\-cifti\-dilate\fR
DILATE A CIFTI FILE
.IP
wb_command \fB\-cifti\-dilate\fR
.IP
<cifti\-in> \- the input cifti file
<direction> \- which dimension to dilate along, ROW or COLUMN
<surface\-distance> \- the distance to dilate on surfaces, in mm
<volume\-distance> \- the distance to dilate in the volume, in mm
<cifti\-out> \- output \- the output cifti file
.IP
[\-left\-surface] \- specify the left surface to use
.IP
<surface> \- the left surface file
.IP
[\-left\-corrected\-areas] \- vertex areas to use instead of computing
.IP
them from the left surface
<area\-metric> \- the corrected vertex areas, as a metric
.IP
[\-right\-surface] \- specify the right surface to use
.IP
<surface> \- the right surface file
.IP
[\-right\-corrected\-areas] \- vertex areas to use instead of computing
.IP
them from the right surface
<area\-metric> \- the corrected vertex areas, as a metric
.IP
[\-cerebellum\-surface] \- specify the cerebellum surface to use
.IP
<surface> \- the cerebellum surface file
.IP
[\-cerebellum\-corrected\-areas] \- vertex areas to use instead of
.IP
computing them from the cerebellum surface
<area\-metric> \- the corrected vertex areas, as a metric
.IP
[\-bad\-brainordinate\-roi] \- specify an roi of brainordinates to overwrite,
.IP
rather than zeros
<roi\-cifti> \- cifti dscalar or dtseries file, positive values denote
.IP
brainordinates to have their values replaced
.IP
[\-nearest] \- use nearest value
.IP
[\-merged\-volume] \- treat volume components as if they were a single
.IP
component
.IP
For all data values designated as bad, if they neighbor a good value or
are within the specified distance of a good value in the same kind of
model, replace the value with a distance weighted average of nearby good
values, otherwise set the value to zero.  If \fB\-nearest\fR is specified, it
will use the value from the closest good value within range instead of a
weighted average.  When the input file contains label data, nearest
dilation is used on the surface, and weighted popularity is used in the
volume.
.IP
The \-*\-corrected\-areas options are intended for dilating on group average
surfaces, but it is only an approximate correction for the reduction of
structure in a group average surface.
.IP
If \fB\-bad\-brainordinate\-roi\fR is specified, all values, including those with
value zero, are good, except for locations with a positive value in the
ROI.  If it is not specified, only values equal to zero are bad.
.PP
\fB\-cifti\-erode\fR
ERODE A CIFTI FILE
.IP
wb_command \fB\-cifti\-erode\fR
.IP
<cifti\-in> \- the input cifti file
<direction> \- which dimension to dilate along, ROW or COLUMN
<surface\-distance> \- the distance to dilate on surfaces, in mm
<volume\-distance> \- the distance to dilate in the volume, in mm
<cifti\-out> \- output \- the output cifti file
.IP
[\-left\-surface] \- specify the left surface to use
.IP
<surface> \- the left surface file
.IP
[\-left\-corrected\-areas] \- vertex areas to use instead of computing
.IP
them from the left surface
<area\-metric> \- the corrected vertex areas, as a metric
.IP
[\-right\-surface] \- specify the right surface to use
.IP
<surface> \- the right surface file
.IP
[\-right\-corrected\-areas] \- vertex areas to use instead of computing
.IP
them from the right surface
<area\-metric> \- the corrected vertex areas, as a metric
.IP
[\-cerebellum\-surface] \- specify the cerebellum surface to use
.IP
<surface> \- the cerebellum surface file
.IP
[\-cerebellum\-corrected\-areas] \- vertex areas to use instead of
.IP
computing them from the cerebellum surface
<area\-metric> \- the corrected vertex areas, as a metric
.IP
[\-merged\-volume] \- treat volume components as if they were a single
.IP
component
.IP
For all data values that are empty (for label data, unlabeled, for other
data, zero), set the surrounding values to empty.  The surrounding values
are defined as the immediate neighbors and all values in the same
structure within the specified distance (\fB\-merged\-volume\fR treats all voxels
as one structure).
.IP
The \-*\-corrected\-areas options are intended for eroding on group average
surfaces, but it is only an approximate correction.
.PP
\fB\-cifti\-estimate\-fwhm\fR
ESTIMATE FWHM SMOOTHNESS OF A CIFTI FILE
.IP
wb_command \fB\-cifti\-estimate\-fwhm\fR
.IP
<cifti> \- the input cifti file
.IP
[\-merged\-volume] \- treat volume components as if they were a single
.IP
component
.IP
[\-column] \- only output estimates for one column
.IP
<column> \- the column number
.IP
[\-surface] \- repeatable \- specify an input surface
.IP
<structure> \- what structure to use this surface for
<surface> \- the surface file
.IP
Estimate the smoothness of the components of the cifti file, printing the
estimates to standard output.  If \fB\-merged\-volume\fR is used, all voxels are
used as a single component, rather than separated by structure.
.IP
<structure> must be one of the following:
.IP
CORTEX_LEFT
CORTEX_RIGHT
CEREBELLUM
ACCUMBENS_LEFT
ACCUMBENS_RIGHT
ALL_GREY_MATTER
ALL_WHITE_MATTER
AMYGDALA_LEFT
AMYGDALA_RIGHT
BRAIN_STEM
CAUDATE_LEFT
CAUDATE_RIGHT
CEREBELLAR_WHITE_MATTER_LEFT
CEREBELLAR_WHITE_MATTER_RIGHT
CEREBELLUM_LEFT
CEREBELLUM_RIGHT
CEREBRAL_WHITE_MATTER_LEFT
CEREBRAL_WHITE_MATTER_RIGHT
CORTEX
DIENCEPHALON_VENTRAL_LEFT
DIENCEPHALON_VENTRAL_RIGHT
HIPPOCAMPUS_LEFT
HIPPOCAMPUS_RIGHT
INVALID
OTHER
OTHER_GREY_MATTER
OTHER_WHITE_MATTER
PALLIDUM_LEFT
PALLIDUM_RIGHT
PUTAMEN_LEFT
PUTAMEN_RIGHT
THALAMUS_LEFT
THALAMUS_RIGHT
.PP
\fB\-cifti\-export\-dense\-mapping\fR
WRITE INDEX TO ELEMENT MAPPING AS TEXT
.IP
wb_command \fB\-cifti\-export\-dense\-mapping\fR
.IP
<cifti> \- the cifti file
<direction> \- which direction to export the mapping from, ROW or COLUMN
.IP
[\-volume\-all] \- export the the mapping of all voxels
.IP
<text\-out> \- output \- the output text file
.IP
[\-no\-cifti\-index] \- don't write the cifti index in the output file
.IP
[\-structure] \- write the structure each voxel belongs to in the output
.IP
file
.IP
[\-surface] \- repeatable \- export the the mapping of one surface structure
.IP
<structure> \- the structure to output
<text\-out> \- output \- the output text file
.IP
[\-no\-cifti\-index] \- don't write the cifti index in the output file
.IP
[\-volume] \- repeatable \- export the the mapping of one volume structure
.IP
<structure> \- the structure to output
<text\-out> \- output \- the output text file
.IP
[\-no\-cifti\-index] \- don't write the cifti index in the output file
.IP
This command produces text files that describe the mapping from cifti
indices to surface vertices or voxels.  All indices are zero\-based.  The
default format for \fB\-surface\fR is lines of the form:
.IP
<cifti\-index> <vertex>
.IP
The default format for \fB\-volume\fR and \fB\-volume\-all\fR is lines of the form:
.IP
<cifti\-index> <i> <j> <k>
.IP
For each <structure> argument, use one of the following strings:
.IP
CORTEX_LEFT
CORTEX_RIGHT
CEREBELLUM
ACCUMBENS_LEFT
ACCUMBENS_RIGHT
ALL_GREY_MATTER
ALL_WHITE_MATTER
AMYGDALA_LEFT
AMYGDALA_RIGHT
BRAIN_STEM
CAUDATE_LEFT
CAUDATE_RIGHT
CEREBELLAR_WHITE_MATTER_LEFT
CEREBELLAR_WHITE_MATTER_RIGHT
CEREBELLUM_LEFT
CEREBELLUM_RIGHT
CEREBRAL_WHITE_MATTER_LEFT
CEREBRAL_WHITE_MATTER_RIGHT
CORTEX
DIENCEPHALON_VENTRAL_LEFT
DIENCEPHALON_VENTRAL_RIGHT
HIPPOCAMPUS_LEFT
HIPPOCAMPUS_RIGHT
INVALID
OTHER
OTHER_GREY_MATTER
OTHER_WHITE_MATTER
PALLIDUM_LEFT
PALLIDUM_RIGHT
PUTAMEN_LEFT
PUTAMEN_RIGHT
THALAMUS_LEFT
THALAMUS_RIGHT
.PP
\fB\-cifti\-extrema\fR
FIND EXTREMA IN A CIFTI FILE
.IP
wb_command \fB\-cifti\-extrema\fR
.IP
<cifti> \- the input cifti
<surface\-distance> \- the minimum distance between extrema of the same
.IP
type, for surface components
.IP
<volume\-distance> \- the minimum distance between extrema of the same
.IP
type, for volume components
.IP
<direction> \- which dimension to find extrema along, ROW or COLUMN
<cifti\-out> \- output \- the output cifti
.IP
[\-left\-surface] \- specify the left surface to use
.IP
<surface> \- the left surface file
.IP
[\-right\-surface] \- specify the right surface to use
.IP
<surface> \- the right surface file
.IP
[\-cerebellum\-surface] \- specify the cerebellum surface to use
.IP
<surface> \- the cerebellum surface file
.IP
[\-surface\-presmooth] \- smooth on the surface before finding extrema
.IP
<surface\-kernel> \- the sigma for the gaussian surface smoothing
.IP
kernel, in mm
.IP
[\-volume\-presmooth] \- smooth volume components before finding extrema
.IP
<volume\-kernel> \- the sigma for the gaussian volume smoothing kernel,
.IP
in mm
.IP
[\-threshold] \- ignore small extrema
.IP
<low> \- the largest value to consider for being a minimum
<high> \- the smallest value to consider for being a maximum
.IP
[\-merged\-volume] \- treat volume components as if they were a single
.IP
component
.IP
[\-sum\-maps] \- output the sum of the extrema maps instead of each map
.IP
separately
.IP
[\-consolidate\-mode] \- use consolidation of local minima instead of a
.IP
large neighborhood
.IP
[\-only\-maxima] \- only find the maxima
.IP
[\-only\-minima] \- only find the minima
.IP
Finds spatial locations in a cifti file that have more extreme values
than all nearby locations in the same component (surface or volume
structure).  The input cifti file must have a brain models mapping along
the specified direction.  COLUMN is the direction that works on dtseries
and dscalar.  For dconn, if it is symmetric use COLUMN, otherwise use
ROW.
.PP
\fB\-cifti\-false\-correlation\fR
COMPARE CORRELATION LOCALLY AND ACROSS/THROUGH SULCI/GYRI
.IP
wb_command \fB\-cifti\-false\-correlation\fR
.IP
<cifti\-in> \- the cifti file to use for correlation
<3D\-dist> \- maximum 3D distance to check around each vertex
<geo\-outer> \- maximum geodesic distance to use for neighboring
.IP
correlation
.IP
<geo\-inner> \- minimum geodesic distance to use for neighboring
.IP
correlation
.IP
<cifti\-out> \- output \- the output cifti dscalar file
.IP
[\-left\-surface] \- specify the left surface to use
.IP
<surface> \- the left surface file
.IP
[\-dump\-text] \- dump the raw measures used to a text file
.IP
<text\-out> \- the output text file
.IP
[\-right\-surface] \- specify the right surface to use
.IP
<surface> \- the right surface file
.IP
[\-dump\-text] \- dump the raw measures used to a text file
.IP
<text\-out> \- the output text file
.IP
[\-cerebellum\-surface] \- specify the cerebellum surface to use
.IP
<surface> \- the cerebellum surface file
.IP
[\-dump\-text] \- dump the raw measures used to a text file
.IP
<text\-out> \- the output text file
.IP
For each vertex, compute the average correlation within a range of
geodesic distances that don't cross a sulcus/gyrus, and the correlation
to the closest vertex crossing a sulcus/gyrus.  A vertex is considered to
cross a sulcus/gyrus if the 3D distance is less than a third of the
geodesic distance.  The output file contains the ratio between these
correlations, and some additional maps to help explain the ratio.
.PP
\fB\-cifti\-find\-clusters\fR
FILTER CLUSTERS BY AREA/VOLUME
.IP
wb_command \fB\-cifti\-find\-clusters\fR
.IP
<cifti> \- the input cifti
<surface\-value\-threshold> \- threshold for surface data values
<surface\-minimum\-area> \- threshold for surface cluster area, in mm^2
<volume\-value\-threshold> \- threshold for volume data values
<volume\-minimum\-size> \- threshold for volume cluster size, in mm^3
<direction> \- which dimension to use for spatial information, ROW or
.IP
COLUMN
.IP
<cifti\-out> \- output \- the output cifti
.IP
[\-less\-than] \- find values less than <value\-threshold>, rather than
.IP
greater
.IP
[\-left\-surface] \- specify the left surface to use
.IP
<surface> \- the left surface file
.IP
[\-corrected\-areas] \- vertex areas to use instead of computing them
.IP
from the surface
<area\-metric> \- the corrected vertex areas, as a metric
.IP
[\-right\-surface] \- specify the right surface to use
.IP
<surface> \- the right surface file
.IP
[\-corrected\-areas] \- vertex areas to use instead of computing them
.IP
from the surface
<area\-metric> \- the corrected vertex areas, as a metric
.IP
[\-cerebellum\-surface] \- specify the cerebellum surface to use
.IP
<surface> \- the cerebellum surface file
.IP
[\-corrected\-areas] \- vertex areas to use instead of computing them
.IP
from the surface
<area\-metric> \- the corrected vertex areas, as a metric
.IP
[\-cifti\-roi] \- search only within regions of interest
.IP
<roi\-cifti> \- the regions to search within, as a cifti file
.IP
[\-merged\-volume] \- treat volume components as if they were a single
.IP
component
.IP
[\-size\-ratio] \- ignore clusters smaller than a given fraction of the
.IP
largest cluster in the structure
<surface\-ratio> \- fraction of the structure's largest cluster area
<volume\-ratio> \- fraction of the structure's largest cluster volume
.IP
[\-distance] \- ignore clusters further than a given distance from the
.IP
largest cluster in the structure
<surface\-distance> \- how far from the largest cluster a cluster can
.IP
be, edge to edge, in mm
.IP
<volume\-distance> \- how far from the largest cluster a cluster can be,
.IP
edge to edge, in mm
.IP
[\-start] \- start labeling clusters from a value other than 1
.IP
<startval> \- the value to give the first cluster found
.IP
Outputs a cifti file with nonzero integers for all brainordinates within
a large enough cluster, and zeros elsewhere.  The integers denote cluster
membership (by default, first cluster found will use value 1, second
cluster 2, etc).  The input cifti file must have a brain models mapping
on the chosen dimension, columns for .dtseries, and either for .dconn.
The ROI should have a brain models mapping along columns, exactly
matching the mapping of the chosen direction in the input file.  Data
outside the ROI is ignored.
.PP
\fB\-cifti\-gradient\fR
TAKE GRADIENT OF A CIFTI FILE
.IP
wb_command \fB\-cifti\-gradient\fR
.IP
<cifti> \- the input cifti
<direction> \- which dimension to take the gradient along, ROW or COLUMN
<cifti\-out> \- output \- the output cifti
.IP
[\-left\-surface] \- specify the left surface to use
.IP
<surface> \- the left surface file
.IP
[\-left\-corrected\-areas] \- vertex areas to use instead of computing
.IP
them from the left surface
<area\-metric> \- the corrected vertex areas, as a metric
.IP
[\-right\-surface] \- specify the right surface to use
.IP
<surface> \- the right surface file
.IP
[\-right\-corrected\-areas] \- vertex areas to use instead of computing
.IP
them from the right surface
<area\-metric> \- the corrected vertex areas, as a metric
.IP
[\-cerebellum\-surface] \- specify the cerebellum surface to use
.IP
<surface> \- the cerebellum surface file
.IP
[\-cerebellum\-corrected\-areas] \- vertex areas to use instead of
.IP
computing them from the cerebellum surface
<area\-metric> \- the corrected vertex areas, as a metric
.IP
[\-surface\-presmooth] \- smooth on the surface before computing the
.IP
gradient
<surface\-kernel> \- the sigma for the gaussian surface smoothing
.IP
kernel, in mm
.IP
[\-volume\-presmooth] \- smooth on the surface before computing the gradient
.IP
<volume\-kernel> \- the sigma for the gaussian volume smoothing kernel,
.IP
in mm
.IP
[\-average\-output] \- output the average of the gradient magnitude maps
.IP
instead of each gradient map separately
.IP
[\-vectors] \- output gradient vectors
.IP
<vectors\-out> \- output \- the vectors, as a dscalar file
.IP
Performs gradient calculation on each component of the cifti file, and
optionally averages the resulting gradients.  The \fB\-vectors\fR and
\fB\-average\-output\fR options may not be used together.  You must specify a
surface for each surface structure in the cifti file.  The COLUMN
direction should be faster, and is the direction that works on dtseries.
For dconn, you probably want ROW, unless you are using \fB\-average\-output\fR.
.PP
\fB\-cifti\-label\-adjacency\fR
MAKE ADJACENCY MATRIX OF A CIFTI LABEL FILE
.IP
wb_command \fB\-cifti\-label\-adjacency\fR
.IP
<label\-in> \- the input cifti label file
<adjacency\-out> \- output \- the output cifti pconn adjacency matrix
.IP
[\-left\-surface] \- specify the left surface to use
.IP
<surface> \- the left surface file
.IP
[\-right\-surface] \- specify the right surface to use
.IP
<surface> \- the right surface file
.IP
[\-cerebellum\-surface] \- specify the cerebellum surface to use
.IP
<surface> \- the cerebellum surface file
.IP
Find face\-adjacent voxels and connected vertices that have different
label values, and count them for each pair.  Put the resulting counts
into a parcellated connectivity file, with the diagonal being zero.  This
gives a rough estimate of how long or expansive the border between two
labels is.
.PP
\fB\-cifti\-label\-export\-table\fR
EXPORT LABEL TABLE FROM CIFTI AS TEXT
.IP
wb_command \fB\-cifti\-label\-export\-table\fR
.IP
<label\-in> \- the input cifti label file
<map> \- the number or name of the label map to use
<table\-out> \- output \- the output text file
.IP
Takes the label table from the cifti label map, and writes it to a text
format matching what is expected by \fB\-cifti\-label\-import\fR.
.PP
\fB\-cifti\-label\-import\fR
MAKE A CIFTI LABEL FILE FROM A CIFTI FILE
.IP
wb_command \fB\-cifti\-label\-import\fR
.IP
<input> \- the input cifti file
<label\-list\-file> \- text file containing the values and names for labels
<output> \- output \- the output cifti label file
.IP
[\-discard\-others] \- set any values not mentioned in the label list to the
.IP
??? label
.IP
[\-unlabeled\-value] \- set the value that will be interpreted as unlabeled
.IP
<value> \- the numeric value for unlabeled (default 0)
.IP
[\-drop\-unused\-labels] \- remove any unused label values from the label
.IP
table
.TP
Creates a cifti label file from a cifti file with label\-like values.
You
.IP
may specify the empty string ('' will work on linux/mac) for
<label\-list\-file>, which will be treated as if it is an empty file.  The
label list file must have lines of the following format:
.IP
<labelname>
<value> <red> <green> <blue> <alpha>
.IP
Do not specify the "unlabeled" key in the file, it is assumed that 0
means not labeled unless \fB\-unlabeled\-value\fR is specified.  Label names must
be on a separate line, but may contain spaces or other unusual characters
(but not newline).  Whitespace is trimmed from both ends of the label
name, but is kept if it is in the middle of a label.  The values of red,
green, blue and alpha must be integers from 0 to 255, and will specify
the color the label is drawn as (alpha of 255 means opaque, which is
probably what you want).  By default, it will set new label names with
names of LABEL_# for any values encountered that are not mentioned in the
list file, specify \fB\-discard\-others\fR to instead set these to the
"unlabeled" key.
.PP
\fB\-cifti\-label\-to\-roi\fR
MAKE A CIFTI LABEL INTO AN ROI
.IP
wb_command \fB\-cifti\-label\-to\-roi\fR
.IP
<label\-in> \- the input cifti label file
<scalar\-out> \- output \- the output cifti scalar file
.IP
[\-name] \- select label by name
.IP
<label\-name> \- the label name that you want an roi of
.IP
[\-key] \- select label by key
.IP
<label\-key> \- the label key that you want an roi of
.IP
[\-map] \- select a single label map to use
.IP
<map> \- the map number or name
.IP
For each map in <label\-in>, a map is created in <scalar\-out> where all
locations labeled with <label\-name> or with a key of <label\-key> are
given a value of 1, and all other locations are given 0.  Exactly one of
\fB\-name\fR and \fB\-key\fR must be specified.  Specify \fB\-map\fR to use only one map from
<label\-in>.
.PP
\fB\-cifti\-math\fR
EVALUATE EXPRESSION ON CIFTI FILES
.IP
wb_command \fB\-cifti\-math\fR
.IP
<expression> \- the expression to evaluate, in quotes
<cifti\-out> \- output \- the output cifti file
.IP
[\-fixnan] \- replace NaN results with a value
.IP
<replace> \- value to replace NaN with
.IP
[\-override\-mapping\-check] \- don't check the mappings for compatibility,
.IP
only check length
.IP
[\-var] \- repeatable \- a cifti file to use as a variable
.IP
<name> \- the name of the variable, as used in the expression
<cifti> \- the cifti file to use as this variable
.IP
[\-select] \- repeatable \- select a single index from a dimension
.IP
<dim> \- the dimension to select from (1\-based)
<index> \- the index to use (1\-based)
.IP
[\-repeat] \- repeat the selected values for each index of output in
.IP
this dimension
.IP
This command evaluates <expression> at each matrix element independently.
There must be at least one \fB\-var\fR option (to get the output layout from),
even if the <name> specified in it isn't used in <expression>.
.IP
To select a single column from a 2D file (most cifti files are 2D), use
\fB\-select\fR 1 <index>, where <index> is 1\-based.  To select a single row from
a 2D file, use \fB\-select\fR 2 <index>.  Where \fB\-select\fR is not used, the cifti
files must have compatible mappings (e.g., brain models and parcels
mappings must match exactly except for parcel names).  Use
\fB\-override\-mapping\-check\fR to skip this checking.
.IP
Filenames are not valid in <expression>, use a variable name and a \fB\-var\fR
option with matching <name> to specify an input file.  The format of
<expression> is as follows:
.IP
Expressions consist of constants, variables, operators, parentheses, and
functions, in infix notation, such as 'exp(\fB\-x\fR + 3) * scale'.  Variables
are strings of any length, using the characters a\-z, A\-Z, 0\-9, and _, but
may not take the name of a named constant.  Currently, there is only one
named constant, PI.  The operators are +, \-, *, /, ^, >, <, >=, <=, ==,
!=, !, &&, ||.  These behave as in C, except that ^ is exponentiation,
i.e. pow(x, y), and takes higher precedence than other binary operators
(also, '\-3^\-4^\-5' means '\-(3^(\-(4^\-5)))').  The <=, >=, ==, and !=
operators are given a small amount of wiggle room, equal to one millionth
of the smaller of the absolute values of the values being compared.
.IP
Comparison and logical operators return 0 or 1, you can do masking with
expressions like 'x * (mask > 0)'.  For all logical operators, an input
is considered true iff it is greater than 0.  The expression '0 < x < 5'
is not syntactically wrong, but it will NOT do what is desired, because
it is evaluated left to right, i.e. '((0 < x) < 5)', which will always
return 1, as both possible results of a comparison are less than 5.  A
warning is generated if an expression of this type is detected.  Use
something like 'x > 0 && x < 5' to get the desired behavior.
.IP
Whitespace between elements is ignored, ' sin ( 2 * x ) ' is equivalent
to 'sin(2*x)', but 's in(2*x)' is an error.  Implied multiplication is
not allowed, the expression '2x' will be parsed as a variable.
Parentheses are (), do not use [] or {}.  Functions require parentheses,
the expression 'sin x' is an error.
.IP
The following functions are supported:
.IP
sin: 1 argument, the sine of the argument (units are radians)
cos: 1 argument, the cosine of the argument (units are radians)
tan: 1 argument, the tangent of the argument (units are radians)
asin: 1 argument, the inverse of sine of the argument, in radians
acos: 1 argument, the inverse of cosine of the argument, in radians
atan: 1 argument, the inverse of tangent of the argument, in radians
atan2: 2 arguments, atan2(y, x) returns the inverse of tangent of
.IP
(y/x), in radians, determining quadrant by the sign of both arguments
.IP
sinh: 1 argument, the hyperbolic sine of the argument
cosh: 1 argument, the hyperbolic cosine of the argument
tanh: 1 argument, the hyperboloc tangent of the argument
asinh: 1 argument, the inverse hyperbolic sine of the argument
acosh: 1 argument, the inverse hyperbolic cosine of the argument
atanh: 1 argument, the inverse hyperboloc tangent of the argument
ln: 1 argument, the natural logarithm of the argument
exp: 1 argument, the constant e raised to the power of the argument
log: 1 argument, the base 10 logarithm of the argument
sqrt: 1 argument, the square root of the argument
abs: 1 argument, the absolute value of the argument
floor: 1 argument, the largest integer not greater than the argument
round: 1 argument, the nearest integer, with ties rounded away from
.IP
zero
.IP
ceil: 1 argument, the smallest integer not less than the argument
min: 2 arguments, min(x, y) returns y if (x > y), x otherwise
max: 2 arguments, max(x, y) returns y if (x < y), x otherwise
mod: 2 arguments, mod(x, y) = x \- y * floor(x / y), or 0 if y == 0
clamp: 3 arguments, clamp(x, low, high) = min(max(x, low), high)
.PP
\fB\-cifti\-merge\fR
MERGE CIFTI TIMESERIES, SCALAR, OR LABEL FILES
.IP
wb_command \fB\-cifti\-merge\fR
.IP
<cifti\-out> \- output \- output cifti file
.IP
[\-cifti] \- repeatable \- specify an input cifti file
.IP
<cifti\-in> \- a cifti file to use columns from
.IP
[\-column] \- repeatable \- select a single column to use
.IP
<column> \- the column index (starting from 1)
.IP
[\-up\-to] \- use an inclusive range of columns
.IP
<last\-column> \- the index of the last column to include
.IP
[\-reverse] \- use the range in reverse order
.IP
Given input CIFTI files which have matching mappings along columns, and
for which mappings along rows are the same type, all either series,
scalars, or labels, this command concatenates the specified columns
horizontally (rows become longer).
.IP
Example: wb_command \fB\-cifti\-merge\fR out.dtseries.nii \fB\-cifti\fR
first.dtseries.nii \fB\-column\fR 1 \fB\-cifti\fR second.dtseries.nii
.IP
This example would take the first column from first.dtseries.nii,
followed by all columns from second.dtseries.nii, and write these columns
to out.dtseries.nii.
.PP
\fB\-cifti\-merge\-dense\fR
MERGE CIFTI FILES ALONG DENSE DIMENSION
.IP
wb_command \fB\-cifti\-merge\-dense\fR
.IP
<direction> \- which dimension to merge along, ROW or COLUMN
<cifti\-out> \- output \- the output cifti file
.IP
[\-cifti] \- repeatable \- specify an input cifti file
.IP
<cifti\-in> \- a cifti file to merge
.IP
The input cifti files must have matching mappings along the direction not
specified, and the mapping along the specified direction must be brain
models.
.PP
\fB\-cifti\-merge\-parcels\fR
MERGE CIFTI FILES ALONG PARCELS DIMENSION
.IP
wb_command \fB\-cifti\-merge\-parcels\fR
.IP
<direction> \- which dimension to merge along (integer, 'ROW', or
.IP
\&'COLUMN')
.IP
<cifti\-out> \- output \- the output cifti file
.IP
[\-cifti] \- repeatable \- specify an input cifti file
.IP
<cifti\-in> \- a cifti file to merge
.IP
The input cifti files must have matching mappings along the direction not
specified, and the mapping along the specified direction must be parcels.
The direction can be either an integer starting from 1, or the strings
\&'ROW' or 'COLUMN'.
.PP
\fB\-cifti\-pairwise\-correlation\fR
CORRELATE PAIRED ROWS BETWEEN TWO CIFTI FILES
.IP
wb_command \fB\-cifti\-pairwise\-correlation\fR
.IP
<cifti\-a> \- first input cifti file
<cifti\-b> \- second input cifti file
<cifti\-out> \- output \- output cifti file
.IP
[\-fisher\-z] \- apply fisher small z transform (ie, artanh) to correlation
.IP
[\-override\-mapping\-check] \- don't check the mappings for compatibility,
.IP
only check length
.IP
For each row in <cifti\-a>, correlate it with the same row in <cifti\-b>,
and put the result in the same row of <cifti\-out>, which has only one
column.
.PP
\fB\-cifti\-palette\fR
SET PALETTE ON A CIFTI FILE
.IP
wb_command \fB\-cifti\-palette\fR
.IP
<cifti\-in> \- the cifti input
<mode> \- the mapping mode
<cifti\-out> \- output \- the output cifti file
.IP
[\-column] \- select a single column for scalar maps
.IP
<column> \- the column number or name
.IP
[\-pos\-percent] \- percentage min/max for positive data coloring
.IP
<pos\-min\-%> \- the percentile for the least positive data
<pos\-max\-%> \- the percentile for the most positive data
.IP
[\-neg\-percent] \- percentage min/max for negative data coloring
.IP
<neg\-min\-%> \- the percentile for the least negative data
<neg\-max\-%> \- the percentile for the most negative data
.IP
[\-pos\-user] \- user min/max values for positive data coloring
.IP
<pos\-min\-user> \- the value for the least positive data
<pos\-max\-user> \- the value for the most positive data
.IP
[\-neg\-user] \- user min/max values for negative data coloring
.IP
<neg\-min\-user> \- the value for the least negative data
<neg\-max\-user> \- the value for the most negative data
.IP
[\-interpolate] \- interpolate colors
.IP
<interpolate> \- boolean, whether to interpolate
.IP
[\-disp\-pos] \- display positive data
.IP
<display> \- boolean, whether to display
.IP
[\-disp\-neg] \- display positive data
.IP
<display> \- boolean, whether to display
.IP
[\-disp\-zero] \- display data closer to zero than the min cutoff
.IP
<display> \- boolean, whether to display
.IP
[\-palette\-name] \- set the palette used
.IP
<name> \- the name of the palette
.IP
[\-thresholding] \- set the thresholding
.IP
<type> \- thresholding setting
<test> \- show values inside or outside thresholds
<min> \- lower threshold
<max> \- upper threshold
.IP
NOTE: The output file must be a different file than the input file.
.IP
For scalar maps, by default the palette is changed for every map, specify
\fB\-column\fR to change only one map.  Palette settings not specified will be
taken from the first column for scalar maps, and from the existing file
palette for other mapping types.  The <mode> argument must be one of the
following:
.IP
MODE_AUTO_SCALE
MODE_AUTO_SCALE_ABSOLUTE_PERCENTAGE
MODE_AUTO_SCALE_PERCENTAGE
MODE_USER_SCALE
.IP
The <name> argument to \fB\-palette\-name\fR must be one of the following:
.IP
PSYCH
PSYCH\-NO\-NONE
ROY\-BIG
ROY\-BIG\-BL
Orange\-Yellow
Gray_Interp_Positive
Gray_Interp
clear_brain
videen_style
fidl
raich4_clrmid
raich6_clrmid
HSB8_clrmid
RBGYR20
RBGYR20P
POS_NEG
red\-yellow
blue\-lightblue
FSL
power_surf
fsl_red
fsl_green
fsl_blue
fsl_yellow
JET256
.IP
The <type> argument to \fB\-thresholding\fR must be one of the following:
.IP
THRESHOLD_TYPE_OFF
THRESHOLD_TYPE_NORMAL
.IP
The <test> argument to \fB\-thresholding\fR must be one of the following:
.IP
THRESHOLD_TEST_SHOW_OUTSIDE
THRESHOLD_TEST_SHOW_INSIDE
.PP
\fB\-cifti\-parcel\-mapping\-to\-label\fR
CREATE DLABEL FROM PARCELLATED FILE
.IP
wb_command \fB\-cifti\-parcel\-mapping\-to\-label\fR
.IP
<cifti\-in> \- the input parcellated file
<direction> \- which dimension to take the parcel map from, ROW or COLUMN
<template\-cifti> \- a cifti file with the desired dense mapping along
.IP
column
.IP
<dlabel\-out> \- output \- the output dense label file
.IP
This command will output a dlabel file, useful for doing the same
parcellation to another dense file.
.IP
For ptseries, pscalar, plabel, pconn, and pdconn, using ROW for
<direction> will work.
.PP
\fB\-cifti\-parcellate\fR
PARCELLATE A CIFTI FILE
.IP
wb_command \fB\-cifti\-parcellate\fR
.IP
<cifti\-in> \- the cifti file to parcellate
<cifti\-label> \- a cifti label file to use for the parcellation
<direction> \- which mapping to parcellate (integer, ROW, or COLUMN)
<cifti\-out> \- output \- output cifti file
.IP
[\-spatial\-weights] \- use voxel volume and either vertex areas or metric
.IP
files as weights
.IP
[\-left\-area\-surf] \- use a surface for left vertex areas
.IP
<left\-surf> \- the left surface to use, areas are in mm^2
.IP
[\-right\-area\-surf] \- use a surface for right vertex areas
.IP
<right\-surf> \- the right surface to use, areas are in mm^2
.IP
[\-cerebellum\-area\-surf] \- use a surface for cerebellum vertex areas
.IP
<cerebellum\-surf> \- the cerebellum surface to use, areas are in
.IP
mm^2
.IP
[\-left\-area\-metric] \- use a metric file for left vertex weights
.IP
<left\-metric> \- metric file containing left vertex weights
.IP
[\-right\-area\-metric] \- use a metric file for right vertex weights
.IP
<right\-metric> \- metric file containing right vertex weights
.IP
[\-cerebellum\-area\-metric] \- use a metric file for cerebellum vertex
.IP
weights
<cerebellum\-metric> \- metric file containing cerebellum vertex
.IP
weights
.IP
[\-cifti\-weights] \- use a cifti file containing weights
.IP
<weight\-cifti> \- the weights to use, as a cifti file
.IP
[\-method] \- specify method of parcellation (default MEAN, or MODE if
.IP
label data)
<method> \- the method to use to assign parcel values from the values
.IP
of member brainordinates
.IP
[\-exclude\-outliers] \- exclude non\-numeric values and outliers from each
.IP
parcel by standard deviation
<sigma\-below> \- number of standard deviations below the mean to
.IP
include
.IP
<sigma\-above> \- number of standard deviations above the mean to
.IP
include
.IP
[\-only\-numeric] \- exclude non\-numeric values
.IP
Each label in the cifti label file will be treated as a parcel, and all
rows or columns within the parcel are averaged together to form the
output row or column.  The direction can be either an integer starting
from 1, or the strings 'ROW' or 'COLUMN'.  For dtseries or dscalar, use
COLUMN.  If you are parcellating a dconn in both directions, parcellating
by ROW first will use much less memory.
.IP
The parameter to the \fB\-method\fR option must be one of the following:
.IP
MAX: the maximum value
MIN: the minimum value
INDEXMAX: the 1\-based index of the maximum value
INDEXMIN: the 1\-based index of the minimum value
SUM: add all values
PRODUCT: multiply all values
MEAN: the mean of the data
STDEV: the standard deviation (N denominator)
SAMPSTDEV: the sample standard deviation (N\-1 denominator)
VARIANCE: the variance of the data
TSNR: mean divided by sample standard deviation (N\-1 denominator)
COV: sample standard deviation (N\-1 denominator) divided by mean
MEDIAN: the median of the data
MODE: the mode of the data
COUNT_NONZERO: the number of nonzero elements in the data
.IP
The \-*\-weights options are mutually exclusive and may only be used with
MEAN, SUM, STDEV, SAMPSTDEV, VARIANCE, MEDIAN, or MODE.
.PP
\fB\-cifti\-reduce\fR
PERFORM REDUCTION OPERATION ON A CIFTI FILE
.IP
wb_command \fB\-cifti\-reduce\fR
.IP
<cifti\-in> \- the cifti file to reduce
<operation> \- the reduction operator to use
<cifti\-out> \- output \- the output cifti file
.IP
[\-direction] \- specify what direction to reduce along
.IP
<direction> \- the direction (default ROW)
.IP
[\-exclude\-outliers] \- exclude non\-numeric values and outliers by standard
.IP
deviation
<sigma\-below> \- number of standard deviations below the mean to
.IP
include
.IP
<sigma\-above> \- number of standard deviations above the mean to
.IP
include
.IP
[\-only\-numeric] \- exclude non\-numeric values
.IP
For the specified direction (default ROW), perform a reduction operation
along that direction.  The direction can be either an integer starting
from 1, or the strings 'ROW' or 'COLUMN'.  The reduction operators are as
follows:
.IP
MAX: the maximum value
MIN: the minimum value
INDEXMAX: the 1\-based index of the maximum value
INDEXMIN: the 1\-based index of the minimum value
SUM: add all values
PRODUCT: multiply all values
MEAN: the mean of the data
STDEV: the standard deviation (N denominator)
SAMPSTDEV: the sample standard deviation (N\-1 denominator)
VARIANCE: the variance of the data
TSNR: mean divided by sample standard deviation (N\-1 denominator)
COV: sample standard deviation (N\-1 denominator) divided by mean
MEDIAN: the median of the data
MODE: the mode of the data
COUNT_NONZERO: the number of nonzero elements in the data
.PP
\fB\-cifti\-reorder\fR
REORDER THE PARCELS OR SCALAR/LABEL MAPS IN A CIFTI FILE
.IP
wb_command \fB\-cifti\-reorder\fR
.IP
<cifti\-in> \- input cifti file
<direction> \- which dimension to reorder along, ROW or COLUMN
<reorder\-list> \- a text file containing the desired order transformation
<cifti\-out> \- output \- the reordered cifti file
.IP
The mapping along the specified direction must be parcels, scalars, or
labels.  For pscalar or ptseries, use COLUMN to reorder the parcels.  For
dlabel, use ROW.  The <reorder\-list> file must contain 1\-based indices
separated by whitespace (spaces, newlines, tabs, etc), with as many
indices as <cifti\-in> has along the specified dimension.  These indices
specify which current index should end up in that position, for instance,
if the current order is 'A B C D', and the desired order is 'D A B C',
the text file should contain '4 1 2 3'.
.PP
\fB\-cifti\-replace\-structure\fR
REPLACE DATA IN A STRUCTURE IN A CIFTI FILE
.IP
wb_command \fB\-cifti\-replace\-structure\fR
.IP
<cifti> \- the cifti to modify
<direction> \- which dimension to interpret as a single map, ROW or COLUMN
.IP
[\-volume\-all] \- replace the data in all volume components
.IP
<volume> \- the input volume
.IP
[\-from\-cropped] \- the input is cropped to the size of the data
.IP
[\-discard\-unused\-labels] \- when operating on a dlabel file, drop any
.IP
unused label keys from the label table
.IP
[\-label] \- repeatable \- replace the data in a surface label component
.IP
<structure> \- the structure to replace the data of
<label> \- the input label file
.IP
[\-metric] \- repeatable \- replace the data in a surface component
.IP
<structure> \- the structure to replace the data of
<metric> \- the input metric
.IP
[\-volume] \- repeatable \- replace the data in a volume component
.IP
<structure> \- the structure to replace the data of
<volume> \- the input volume
.IP
[\-from\-cropped] \- the input is cropped to the size of the component
.IP
You must specify at least one of \fB\-metric\fR, \fB\-label\fR, \fB\-volume\fR, or \fB\-volume\-all\fR
for this command to do anything.  Input volumes must line up with the
output of \fB\-cifti\-separate\fR.  For dtseries/dscalar, use COLUMN, and if your
matrix will be fully symmetric, COLUMN is more efficient.  The structure
argument must be one of the following:
.IP
CORTEX_LEFT
CORTEX_RIGHT
CEREBELLUM
ACCUMBENS_LEFT
ACCUMBENS_RIGHT
ALL_GREY_MATTER
ALL_WHITE_MATTER
AMYGDALA_LEFT
AMYGDALA_RIGHT
BRAIN_STEM
CAUDATE_LEFT
CAUDATE_RIGHT
CEREBELLAR_WHITE_MATTER_LEFT
CEREBELLAR_WHITE_MATTER_RIGHT
CEREBELLUM_LEFT
CEREBELLUM_RIGHT
CEREBRAL_WHITE_MATTER_LEFT
CEREBRAL_WHITE_MATTER_RIGHT
CORTEX
DIENCEPHALON_VENTRAL_LEFT
DIENCEPHALON_VENTRAL_RIGHT
HIPPOCAMPUS_LEFT
HIPPOCAMPUS_RIGHT
INVALID
OTHER
OTHER_GREY_MATTER
OTHER_WHITE_MATTER
PALLIDUM_LEFT
PALLIDUM_RIGHT
PUTAMEN_LEFT
PUTAMEN_RIGHT
THALAMUS_LEFT
THALAMUS_RIGHT
.PP
\fB\-cifti\-resample\fR
RESAMPLE A CIFTI FILE TO A NEW CIFTI SPACE
.IP
wb_command \fB\-cifti\-resample\fR
.IP
<cifti\-in> \- the cifti file to resample
<direction> \- the direction of the input that should be resampled
<cifti\-template> \- a cifti file containing the cifti space to resample to
<template\-direction> \- the direction of the template to use as the
.IP
resampling space
.IP
<surface\-method> \- specify a surface resampling method
<volume\-method> \- specify a volume interpolation method
<cifti\-out> \- output \- the output cifti file
.IP
[\-surface\-largest] \- use largest weight instead of weighted average or
.IP
popularity when doing surface resampling
.IP
[\-volume\-predilate] \- dilate the volume components before resampling
.IP
<dilate\-mm> \- distance, in mm, to dilate
.IP
[\-nearest] \- use nearest value dilation
.IP
[\-weighted] \- use weighted dilation (default)
.IP
[\-exponent] \- specify exponent in weighting function
.IP
<exponent> \- exponent 'n' to use in (1 / (distance ^ n)) as the
.IP
weighting function (default 2)
.IP
[\-surface\-postdilate] \- dilate the surface components after resampling
.IP
<dilate\-mm> \- distance, in mm, to dilate
.IP
[\-nearest] \- use nearest value dilation
.IP
[\-linear] \- use linear dilation
.IP
[\-weighted] \- use weighted dilation (default for non\-label data)
.IP
[\-exponent] \- specify exponent in weighting function
.IP
<exponent> \- exponent 'n' to use in (area / (distance ^ n)) as
.IP
the weighting function (default 2)
.IP
[\-affine] \- use an affine transformation on the volume components
.IP
<affine\-file> \- the affine file to use
.IP
[\-flirt] \- MUST be used if affine is a flirt affine
.IP
<source\-volume> \- the source volume used when generating the affine
<target\-volume> \- the target volume used when generating the affine
.IP
[\-warpfield] \- use a warpfield on the volume components
.IP
<warpfield> \- the warpfield to use
.IP
[\-fnirt] \- MUST be used if using a fnirt warpfield
.IP
<source\-volume> \- the source volume used when generating the
.IP
warpfield
.IP
[\-left\-spheres] \- specify spheres for left surface resampling
.IP
<current\-sphere> \- a sphere with the same mesh as the current left
.IP
surface
.IP
<new\-sphere> \- a sphere with the new left mesh that is in register
.IP
with the current sphere
.IP
[\-left\-area\-surfs] \- specify left surfaces to do vertex area
.IP
correction based on
<current\-area> \- a relevant left anatomical surface with current
.IP
mesh
.IP
<new\-area> \- a relevant left anatomical surface with new mesh
.IP
[\-left\-area\-metrics] \- specify left vertex area metrics to do area
.IP
correction based on
<current\-area> \- a metric file with vertex areas for the current
.IP
mesh
.IP
<new\-area> \- a metric file with vertex areas for the new mesh
.IP
[\-right\-spheres] \- specify spheres for right surface resampling
.IP
<current\-sphere> \- a sphere with the same mesh as the current right
.IP
surface
.IP
<new\-sphere> \- a sphere with the new right mesh that is in register
.IP
with the current sphere
.IP
[\-right\-area\-surfs] \- specify right surfaces to do vertex area
.IP
correction based on
<current\-area> \- a relevant right anatomical surface with current
.IP
mesh
.IP
<new\-area> \- a relevant right anatomical surface with new mesh
.IP
[\-right\-area\-metrics] \- specify right vertex area metrics to do area
.IP
correction based on
<current\-area> \- a metric file with vertex areas for the current
.IP
mesh
.IP
<new\-area> \- a metric file with vertex areas for the new mesh
.IP
[\-cerebellum\-spheres] \- specify spheres for cerebellum surface resampling
.IP
<current\-sphere> \- a sphere with the same mesh as the current
.IP
cerebellum surface
.IP
<new\-sphere> \- a sphere with the new cerebellum mesh that is in
.IP
register with the current sphere
.IP
[\-cerebellum\-area\-surfs] \- specify cerebellum surfaces to do vertex
.IP
area correction based on
<current\-area> \- a relevant cerebellum anatomical surface with
.IP
current mesh
.IP
<new\-area> \- a relevant cerebellum anatomical surface with new mesh
.IP
[\-cerebellum\-area\-metrics] \- specify cerebellum vertex area metrics to
.IP
do area correction based on
<current\-area> \- a metric file with vertex areas for the current
.IP
mesh
.IP
<new\-area> \- a metric file with vertex areas for the new mesh
.TP
Resample cifti data to a different brainordinate space.
Use COLUMN for
.TP
the direction to resample dscalar, dlabel, or dtseries.
Resampling both
.IP
dimensions of a dconn requires running this command twice, once with
COLUMN and once with ROW.  If you are resampling a dconn and your machine
has a large amount of memory, you might consider using
\fB\-cifti\-resample\-dconn\-memory\fR to avoid writing and rereading an
intermediate file.  If spheres are not specified for a surface structure
which exists in the cifti files, its data is copied without resampling or
dilation.  Dilation is done with the 'nearest' method, and is done on
<new\-sphere> for surface data.  Volume components are padded before
dilation so that dilation doesn't run into the edge of the component
bounding box.  If neither \fB\-affine\fR nor \fB\-warpfield\fR are specified, the
identity transform is assumed for the volume data.
.IP
The recommended resampling methods are ADAP_BARY_AREA and CUBIC (cubic
spline), except for label data which should use ADAP_BARY_AREA and
ENCLOSING_VOXEL.  Using ADAP_BARY_AREA requires specifying an area option
to each used \-*\-spheres option.
.IP
The <volume\-method> argument must be one of the following:
.IP
CUBIC
ENCLOSING_VOXEL
TRILINEAR
.IP
The <surface\-method> argument must be one of the following:
.IP
ADAP_BARY_AREA
BARYCENTRIC
.PP
\fB\-cifti\-resample\-dconn\-memory\fR
USE LOTS OF MEMORY TO RESAMPLE DCONN
.IP
wb_command \fB\-cifti\-resample\-dconn\-memory\fR
.IP
<cifti\-in> \- the cifti file to resample
<cifti\-template> \- a cifti file containing the cifti space to resample to
<template\-direction> \- the direction of the template to use as the
.IP
resampling space
.IP
<surface\-method> \- specify a surface resampling method
<volume\-method> \- specify a volume interpolation method
<cifti\-out> \- output \- the output cifti file
.IP
[\-surface\-largest] \- use largest weight instead of weighted average when
.IP
doing surface resampling
.IP
[\-volume\-predilate] \- dilate the volume components before resampling
.IP
<dilate\-mm> \- distance, in mm, to dilate
.IP
[\-nearest] \- use nearest value dilation
.IP
[\-weighted] \- use weighted dilation
.IP
[\-exponent] \- specify exponent in weighting function
.IP
<exponent> \- exponent 'n' to use in (1 / (distance ^ n)) as the
.IP
weighting function (default 2)
.IP
[\-surface\-postdilate] \- dilate the surface components after resampling
.IP
<dilate\-mm> \- distance, in mm, to dilate
.IP
[\-nearest] \- use nearest value dilation
.IP
[\-linear] \- use linear dilation
.IP
[\-weighted] \- use weighted dilation
.IP
[\-exponent] \- specify exponent in weighting function
.IP
<exponent> \- exponent 'n' to use in (area / (distance ^ n)) as
.IP
the weighting function (default 2)
.IP
[\-affine] \- use an affine transformation on the volume components
.IP
<affine\-file> \- the affine file to use
.IP
[\-flirt] \- MUST be used if affine is a flirt affine
.IP
<source\-volume> \- the source volume used when generating the affine
<target\-volume> \- the target volume used when generating the affine
.IP
[\-warpfield] \- use a warpfield on the volume components
.IP
<warpfield> \- the warpfield to use
.IP
[\-fnirt] \- MUST be used if using a fnirt warpfield
.IP
<source\-volume> \- the source volume used when generating the
.IP
warpfield
.IP
[\-left\-spheres] \- specify spheres for left surface resampling
.IP
<current\-sphere> \- a sphere with the same mesh as the current left
.IP
surface
.IP
<new\-sphere> \- a sphere with the new left mesh that is in register
.IP
with the current sphere
.IP
[\-left\-area\-surfs] \- specify left surfaces to do vertex area
.IP
correction based on
<current\-area> \- a relevant left anatomical surface with current
.IP
mesh
.IP
<new\-area> \- a relevant left anatomical surface with new mesh
.IP
[\-left\-area\-metrics] \- specify left vertex area metrics to do area
.IP
correction based on
<current\-area> \- a metric file with vertex areas for the current
.IP
mesh
.IP
<new\-area> \- a metric file with vertex areas for the new mesh
.IP
[\-right\-spheres] \- specify spheres for right surface resampling
.IP
<current\-sphere> \- a sphere with the same mesh as the current right
.IP
surface
.IP
<new\-sphere> \- a sphere with the new right mesh that is in register
.IP
with the current sphere
.IP
[\-right\-area\-surfs] \- specify right surfaces to do vertex area
.IP
correction based on
<current\-area> \- a relevant right anatomical surface with current
.IP
mesh
.IP
<new\-area> \- a relevant right anatomical surface with new mesh
.IP
[\-right\-area\-metrics] \- specify right vertex area metrics to do area
.IP
correction based on
<current\-area> \- a metric file with vertex areas for the current
.IP
mesh
.IP
<new\-area> \- a metric file with vertex areas for the new mesh
.IP
[\-cerebellum\-spheres] \- specify spheres for cerebellum surface resampling
.IP
<current\-sphere> \- a sphere with the same mesh as the current
.IP
cerebellum surface
.IP
<new\-sphere> \- a sphere with the new cerebellum mesh that is in
.IP
register with the current sphere
.IP
[\-cerebellum\-area\-surfs] \- specify cerebellum surfaces to do vertex
.IP
area correction based on
<current\-area> \- a relevant cerebellum anatomical surface with
.IP
current mesh
.IP
<new\-area> \- a relevant cerebellum anatomical surface with new mesh
.IP
[\-cerebellum\-area\-metrics] \- specify cerebellum vertex area metrics to
.IP
do area correction based on
<current\-area> \- a metric file with vertex areas for the current
.IP
mesh
.IP
<new\-area> \- a metric file with vertex areas for the new mesh
.IP
This command does the same thing as running \fB\-cifti\-resample\fR twice, but
uses memory up to approximately 2x the size that the intermediate file
would be.  This is because the intermediate dconn is kept in memory,
rather than written to disk, and the components before and after
resampling/dilation have to be in memory at the same time during the
relevant computation.  If spheres are not specified for a surface
structure which exists in the cifti files, its data is copied without
resampling or dilation.  Dilation is done with the 'nearest' method, and
is done on <new\-sphere> for surface data.  Volume components are padded
before dilation so that dilation doesn't run into the edge of the
component bounding box.
.IP
The <volume\-method> argument must be one of the following:
.IP
CUBIC
ENCLOSING_VOXEL
TRILINEAR
.IP
The <surface\-method> argument must be one of the following:
.IP
ADAP_BARY_AREA
BARYCENTRIC
.PP
\fB\-cifti\-restrict\-dense\-map\fR
EXCLUDE BRAINORDINATES FROM A CIFTI FILE
.IP
wb_command \fB\-cifti\-restrict\-dense\-map\fR
.IP
<cifti\-in> \- the input cifti
<direction> \- which dimension to change the mapping on (integer, 'ROW',
.IP
or 'COLUMN')
.IP
<cifti\-out> \- output \- the output cifti
.IP
[\-cifti\-roi] \- cifti file containing combined rois
.IP
<roi\-cifti> \- the rois as a cifti file
.IP
[\-left\-roi] \- vertices to use from left hemisphere
.IP
<roi\-metric> \- the left roi as a metric file
.IP
[\-right\-roi] \- vertices to use from right hemisphere
.IP
<roi\-metric> \- the right roi as a metric file
.IP
[\-cerebellum\-roi] \- vertices to use from cerebellum
.IP
<roi\-metric> \- the cerebellum roi as a metric file
.IP
[\-vol\-roi] \- voxels to use
.IP
<roi\-vol> \- the roi volume file
.IP
Writes a modified version of <cifti\-in>, where all brainordinates outside
the specified roi(s) are removed from the file.  The direction can be
either an integer starting from 1, or the strings 'ROW' or 'COLUMN'.  If
\fB\-cifti\-roi\fR is specified, no other \-*\-roi option may be specified.  If not
using \fB\-cifti\-roi\fR, any \-*\-roi options not present will discard the
relevant structure, if present in the input file.
.PP
\fB\-cifti\-roi\-average\fR
AVERAGE ROWS IN A SINGLE CIFTI FILE
.IP
wb_command \fB\-cifti\-roi\-average\fR
.IP
<cifti\-in> \- the cifti file to average rows from
<text\-out> \- output text file of the average values
.IP
[\-cifti\-roi] \- cifti file containing combined rois
.IP
<roi\-cifti> \- the rois as a cifti file
.IP
[\-left\-roi] \- vertices to use from left hemisphere
.IP
<roi\-metric> \- the left roi as a metric file
.IP
[\-right\-roi] \- vertices to use from right hemisphere
.IP
<roi\-metric> \- the right roi as a metric file
.IP
[\-cerebellum\-roi] \- vertices to use from cerebellum
.IP
<roi\-metric> \- the cerebellum roi as a metric file
.IP
[\-vol\-roi] \- voxels to use
.IP
<roi\-vol> \- the roi volume file
.IP
Average the rows that are within the specified ROIs, and write the
resulting average row to a text file, separated by newlines.  If
\fB\-cifti\-roi\fR is specified, \fB\-left\-roi\fR, \fB\-right\-roi\fR, \fB\-cerebellum\-roi\fR, and
\fB\-vol\-roi\fR must not be specified.
.PP
\fB\-cifti\-rois\-from\-extrema\fR
CREATE CIFTI ROI MAPS FROM EXTREMA MAPS
.IP
wb_command \fB\-cifti\-rois\-from\-extrema\fR
.IP
<cifti> \- the input cifti
<surf\-limit> \- geodesic distance limit from vertex, in mm
<vol\-limit> \- euclidean distance limit from voxel center, in mm
<direction> \- which dimension an extrema map is along, ROW or COLUMN
<cifti\-out> \- output \- the output cifti
.IP
[\-left\-surface] \- specify the left surface to use
.IP
<surface> \- the left surface file
.IP
[\-right\-surface] \- specify the right surface to use
.IP
<surface> \- the right surface file
.IP
[\-cerebellum\-surface] \- specify the cerebellum surface to use
.IP
<surface> \- the cerebellum surface file
.IP
[\-gaussian] \- generate gaussian kernels instead of flat ROIs
.IP
<surf\-sigma> \- the sigma for the surface gaussian kernel, in mm
<vol\-sigma> \- the sigma for the volume gaussian kernel, in mm
.IP
[\-overlap\-logic] \- how to handle overlapping ROIs, default ALLOW
.IP
<method> \- the method of resolving overlaps
.IP
[\-merged\-volume] \- treat volume components as if they were a single
.IP
component
.IP
For each nonzero value in each map, make a map with an ROI around that
location.  If the \fB\-gaussian\fR option is specified, then normalized gaussian
kernels are output instead of ROIs.  The <method> argument to
\fB\-overlap\-logic\fR must be one of ALLOW, CLOSEST, or EXCLUDE.  ALLOW is the
default, and means that ROIs are treated independently and may overlap.
CLOSEST means that ROIs may not overlap, and that no ROI contains
vertices that are closer to a different seed vertex.  EXCLUDE means that
ROIs may not overlap, and that any vertex within range of more than one
ROI does not belong to any ROI.
.PP
\fB\-cifti\-separate\fR
WRITE A CIFTI STRUCTURE AS METRIC, LABEL OR VOLUME
.IP
wb_command \fB\-cifti\-separate\fR
.IP
<cifti\-in> \- the cifti to separate a component of
<direction> \- which direction to separate into components, ROW or COLUMN
.IP
[\-volume\-all] \- separate all volume structures into a volume file
.IP
<volume\-out> \- output \- the output volume
.IP
[\-roi] \- also output the roi of which voxels have data
.IP
<roi\-out> \- output \- the roi output volume
.IP
[\-label] \- output a volume label file indicating the location of
.IP
structures
<label\-out> \- output \- the label output volume
.IP
[\-crop] \- crop volume to the size of the data rather than using the
.IP
original volume size
.IP
[\-label] \- repeatable \- separate a surface model into a surface label
.IP
file
<structure> \- the structure to output
<label\-out> \- output \- the output label file
.IP
[\-roi] \- also output the roi of which vertices have data
.IP
<roi\-out> \- output \- the roi output metric
.IP
[\-metric] \- repeatable \- separate a surface model into a metric file
.IP
<structure> \- the structure to output
<metric\-out> \- output \- the output metric
.IP
[\-roi] \- also output the roi of which vertices have data
.IP
<roi\-out> \- output \- the roi output metric
.IP
[\-volume] \- repeatable \- separate a volume structure into a volume file
.IP
<structure> \- the structure to output
<volume\-out> \- output \- the output volume
.IP
[\-roi] \- also output the roi of which voxels have data
.IP
<roi\-out> \- output \- the roi output volume
.IP
[\-crop] \- crop volume to the size of the component rather than using
.IP
the original volume size
.IP
For dtseries, dscalar, and dlabel, use COLUMN for <direction>, and if you
have a symmetric dconn, COLUMN is more efficient.
.IP
You must specify at least one of \fB\-metric\fR, \fB\-volume\-all\fR, \fB\-volume\fR, or \fB\-label\fR
for this command to do anything.  Output volumes will spatially line up
with their original positions, whether or not they are cropped.
.IP
For each <structure> argument, use one of the following strings:
.IP
CORTEX_LEFT
CORTEX_RIGHT
CEREBELLUM
ACCUMBENS_LEFT
ACCUMBENS_RIGHT
ALL_GREY_MATTER
ALL_WHITE_MATTER
AMYGDALA_LEFT
AMYGDALA_RIGHT
BRAIN_STEM
CAUDATE_LEFT
CAUDATE_RIGHT
CEREBELLAR_WHITE_MATTER_LEFT
CEREBELLAR_WHITE_MATTER_RIGHT
CEREBELLUM_LEFT
CEREBELLUM_RIGHT
CEREBRAL_WHITE_MATTER_LEFT
CEREBRAL_WHITE_MATTER_RIGHT
CORTEX
DIENCEPHALON_VENTRAL_LEFT
DIENCEPHALON_VENTRAL_RIGHT
HIPPOCAMPUS_LEFT
HIPPOCAMPUS_RIGHT
INVALID
OTHER
OTHER_GREY_MATTER
OTHER_WHITE_MATTER
PALLIDUM_LEFT
PALLIDUM_RIGHT
PUTAMEN_LEFT
PUTAMEN_RIGHT
THALAMUS_LEFT
THALAMUS_RIGHT
.PP
\fB\-cifti\-smoothing\fR
SMOOTH A CIFTI FILE
.IP
wb_command \fB\-cifti\-smoothing\fR
.IP
<cifti> \- the input cifti
<surface\-kernel> \- the sigma for the gaussian surface smoothing kernel,
.IP
in mm
.IP
<volume\-kernel> \- the sigma for the gaussian volume smoothing kernel, in
.IP
mm
.IP
<direction> \- which dimension to smooth along, ROW or COLUMN
<cifti\-out> \- output \- the output cifti
.IP
[\-left\-surface] \- specify the left surface to use
.IP
<surface> \- the left surface file
.IP
[\-left\-corrected\-areas] \- vertex areas to use instead of computing
.IP
them from the left surface
<area\-metric> \- the corrected vertex areas, as a metric
.IP
[\-right\-surface] \- specify the right surface to use
.IP
<surface> \- the right surface file
.IP
[\-right\-corrected\-areas] \- vertex areas to use instead of computing
.IP
them from the right surface
<area\-metric> \- the corrected vertex areas, as a metric
.IP
[\-cerebellum\-surface] \- specify the cerebellum surface to use
.IP
<surface> \- the cerebellum surface file
.IP
[\-cerebellum\-corrected\-areas] \- vertex areas to use instead of
.IP
computing them from the cerebellum surface
<area\-metric> \- the corrected vertex areas, as a metric
.IP
[\-cifti\-roi] \- smooth only within regions of interest
.IP
<roi\-cifti> \- the regions to smooth within, as a cifti file
.IP
[\-fix\-zeros\-volume] \- treat values of zero in the volume as missing data
.IP
[\-fix\-zeros\-surface] \- treat values of zero on the surface as missing
.IP
data
.IP
[\-merged\-volume] \- smooth across subcortical structure boundaries
.IP
The input cifti file must have a brain models mapping on the chosen
dimension, columns for .dtseries, and either for .dconn.  By default,
data in different structures is smoothed independently (i.e., "parcel
constrained" smoothing), so volume structures that touch do not smooth
across this boundary.  Specify \fB\-merged\-volume\fR to ignore these boundaries.
Surface smoothing uses the GEO_GAUSS_AREA smoothing method.
.IP
The \-*\-corrected\-areas options are intended for when it is unavoidable to
smooth on group average surfaces, it is only an approximate correction
for the reduction of structure in a group average surface.  It is better
to smooth the data on individuals before averaging, when feasible.
.IP
The \fB\-fix\-zeros\-\fR* options will treat values of zero as lack of data, and
not use that value when generating the smoothed values, but will fill
zeros with extrapolated values.  The ROI should have a brain models
mapping along columns, exactly matching the mapping of the chosen
direction in the input file.  Data outside the ROI is ignored.
.PP
\fB\-cifti\-stats\fR
STATISTICS ALONG CIFTI COLUMNS
.IP
wb_command \fB\-cifti\-stats\fR
.IP
<cifti\-in> \- the input cifti
.IP
[\-reduce] \- use a reduction operation
.IP
<operation> \- the reduction operation
.IP
[\-percentile] \- give the value at a percentile
.IP
<percent> \- the percentile to find
.IP
[\-column] \- only display output for one column
.IP
<column> \- the column index (starting from 1)
.IP
[\-roi] \- only consider data inside an roi
.IP
<roi\-cifti> \- the roi, as a cifti file
.IP
[\-match\-maps] \- each column of input uses the corresponding column
.IP
from the roi file
.IP
[\-show\-map\-name] \- print column index and name before each output
.IP
For each column of the input, a single number is printed, resulting from
the specified reduction or percentile operation.  Use \fB\-column\fR to only
give output for a single column.  Use \fB\-roi\fR to consider only the data
within a region.  Exactly one of \fB\-reduce\fR or \fB\-percentile\fR must be
specified.
.IP
The argument to the \fB\-reduce\fR option must be one of the following:
.IP
MAX: the maximum value
MIN: the minimum value
INDEXMAX: the 1\-based index of the maximum value
INDEXMIN: the 1\-based index of the minimum value
SUM: add all values
PRODUCT: multiply all values
MEAN: the mean of the data
STDEV: the standard deviation (N denominator)
SAMPSTDEV: the sample standard deviation (N\-1 denominator)
VARIANCE: the variance of the data
TSNR: mean divided by sample standard deviation (N\-1 denominator)
COV: sample standard deviation (N\-1 denominator) divided by mean
MEDIAN: the median of the data
MODE: the mode of the data
COUNT_NONZERO: the number of nonzero elements in the data
.PP
\fB\-cifti\-transpose\fR
TRANSPOSE A CIFTI FILE
.IP
wb_command \fB\-cifti\-transpose\fR
.IP
<cifti\-in> \- the input cifti file
<cifti\-out> \- output \- the output cifti file
.IP
[\-mem\-limit] \- restrict memory usage
.IP
<limit\-GB> \- memory limit in gigabytes
.TP
The input must be a 2\-dimensional cifti file.
The output is a cifti file
.IP
where every row in the input is a column in the output.
.PP
\fB\-cifti\-vector\-operation\fR
DO A VECTOR OPERATION ON CIFTI FILES
.IP
wb_command \fB\-cifti\-vector\-operation\fR
.IP
<vectors\-a> \- first vector input file
<vectors\-b> \- second vector input file
<operation> \- what vector operation to do
<cifti\-out> \- output \- the output file
.IP
[\-normalize\-a] \- normalize vectors of first input
.IP
[\-normalize\-b] \- normalize vectors of second input
.IP
[\-normalize\-output] \- normalize output vectors (not valid for dot
.IP
product)
.IP
[\-magnitude] \- output the magnitude of the result (not valid for dot
.IP
product)
.IP
Does a vector operation on two cifti files (that must have a multiple of
3 columns).  Either of the inputs may have multiple vectors (more than 3
columns), but not both (at least one must have exactly 3 columns).  The
\fB\-magnitude\fR and \fB\-normalize\-output\fR options may not be specified together,
or with an operation that returns a scalar (dot product).  The
<operation> parameter must be one of the following:
.IP
DOT
CROSS
ADD
SUBTRACT
.PP
\fB\-cifti\-weighted\-stats\fR
WEIGHTED STATISTICS ALONG CIFTI COLUMNS
.IP
wb_command \fB\-cifti\-weighted\-stats\fR
.IP
<cifti\-in> \- the input cifti
.IP
[\-spatial\-weights] \- use vertex area and voxel volume as weights
.IP
[\-left\-area\-surf] \- use a surface for left vertex areas
.IP
<left\-surf> \- the left surface to use, areas are in mm^2
.IP
[\-right\-area\-surf] \- use a surface for right vertex areas
.IP
<right\-surf> \- the right surface to use, areas are in mm^2
.IP
[\-cerebellum\-area\-surf] \- use a surface for cerebellum vertex areas
.IP
<cerebellum\-surf> \- the cerebellum surface to use, areas are in
.IP
mm^2
.IP
[\-left\-area\-metric] \- use a metric file for left vertex areas
.IP
<left\-metric> \- metric file containing left vertex areas
.IP
[\-right\-area\-metric] \- use a metric file for right vertex areas
.IP
<right\-metric> \- metric file containing right vertex areas
.IP
[\-cerebellum\-area\-metric] \- use a metric file for cerebellum vertex
.IP
areas
<cerebellum\-metric> \- metric file containing cerebellum vertex
.IP
areas
.IP
[\-cifti\-weights] \- use a cifti file containing weights
.IP
<weight\-cifti> \- the weights to use, as a cifti file
.IP
[\-column] \- only display output for one column
.IP
<column> \- the column to use (1\-based)
.IP
[\-roi] \- only consider data inside an roi
.IP
<roi\-cifti> \- the roi, as a cifti file
.IP
[\-match\-maps] \- each column of input uses the corresponding column
.IP
from the roi file
.IP
[\-mean] \- compute weighted mean
.IP
[\-stdev] \- compute weighted standard deviation
.IP
[\-sample] \- estimate population stdev from the sample
.IP
[\-percentile] \- compute weighted percentile
.IP
<percent> \- the percentile to find
.IP
[\-sum] \- compute weighted sum
.IP
[\-show\-map\-name] \- print map index and name before each output
.IP
If the mapping along column is brain models, for each column of the
input, the specified operation is done on each surface and across all
voxels, and the results are printed separately.  For other mapping types,
the operation is done on each column, and one number per map is printed.
Exactly one of \fB\-spatial\-weights\fR or \fB\-cifti\-weights\fR must be specified.  Use
\fB\-column\fR to only give output for a single column.  Use \fB\-roi\fR to consider
only the data within a region.  Exactly one of \fB\-mean\fR, \fB\-stdev\fR, \fB\-percentile\fR
or \fB\-sum\fR must be specified.
.IP
Using \fB\-sum\fR with \fB\-spatial\-weights\fR (or with \fB\-cifti\-weights\fR and a cifti
containing weights of similar meaning) is equivalent to integrating with
respect to area and volume.  When the input is binary ROIs, this will
therefore output the area or volume of each ROI.
.PP
\fB\-class\-add\-member\fR
.PP
Add members to class header (.h) and implementation (.cxx) files.
.IP
[\-add\-to\-files]
[\-m <member\-name> <data\-type> <description>]...
.IP
If the \fB\-add\-to\-files\fR is not specified, the code for the
header and implementation files is printed to the
terminal.
.IP
If the \fB\-add\-to\-files\fR is specified, the class files are
expected to be in the current directory and named
<class\-name>.h and <class\-name>.cxx.  The header
file must contain this text in its private section:
.IP
// ADD_NEW_MEMBERS_HERE
.IP
The implementation file must contain this text in
its public section:
.IP
// ADD_NEW_METHODS_HERE
.IP
If either of these text string are missing, the code
that would have been added to the file(s) is printed
to the terminal.
.IP
For each member, three text strings separated by a space
must be provided and they are the name of the member
its data type, and a description of the member.  If the
description contains spaces the description must be
enclosed in double quotes ("").
.IP
If the data type begins with a capital letter, it is
assumed to be the name of a class.  In this case, both
const and non\-const getters are created but not setter
is created.  Otherwise, the data type is expected to be
a primitive type and both a getter and a setter are
created.  Note that AString and QString are treated as
primitive types.
.PP
\fB\-class\-create\fR
.PP
Create class header (.h) and implementation (.cxx) files.
.PP
Usage:  <class\-name>
.IP
[\-copy]
[\-event\-class <event\-type\-enum>]
[\-event\-listener]
[\-no\-parent]
[\-parent <parent\-class\-name>]
.SH OPTIONS
.HP
\fB\-copy\fR
.IP
Adds copy constructor and assignment operator
.HP
\fB\-event\-class\fR <event\-type\-enum>
.IP
When creating an Event subclass, using this
option will automatically set the parent
class to Event and place the given event
enumerated type value into the parameter
for the Event class constructor.
.IP
For the <event\-type\-enum> there is no need
to prepend it with "EventTypeEnum::".
.HP
\fB\-event\-listener\fR
.IP
Implement the EventListenerInterface so
that the class may listen for events.
.HP
\fB\-no\-parent\fR
.IP
Created class is not derived from any other
class.  By default, the parent class is
CaretObject.
.HP
\fB\-parent\fR <parent\-class\-name>
.IP
Specify the parent (derived from) class.
By default, the parent class is CaretObject.
.HP
\fB\-scene\fR
.IP
Implement the SceneableInterface so that
instances of the class can be restored from
and saved to scenes.
.HP
\fB\-scene\-sub\-class\fR
.IP
Adds methods that can be called by the superclass so that this sub\-class can save and
restore data to and from scenes.
.IP
This option should only be used when creating
a class whose super class implements the
SceneableInterface
.PP
\fB\-class\-create\-algorithm\fR
.PP
Create Algorithm Class header (.h) and implementation (.cxx) files.
.PP
Usage:  <algorithm\-class\-name>
.IP
<command\-line\-switch>
<short\-description>
.IP
algorithm\-class\-name
.IP
Required name of the algorithm class that MUST start with "Algorithm"
.IP
command\-line\-switch
.IP
Required command line switch for algorithm.
.IP
short\-description
.IP
Required short description within double quotes.
.PP
\fB\-class\-create\-enum\fR
.PP
Create enumerated type header (.h) and implementation (.cxx) files.
.PP
Usage:  <enum\-class\-name>
.IP
<number\-of\-values>
<auto\-number>
.IP
enum\-class\-name
.TP
Name of the enumerated type.
Must end in "Enum"
.IP
number\-of\-values
.IP
Number of values in the enumerated type.
.IP
auto\-number
.IP
Automatically generated integer codes corresponding
to the enumerated values.  Value for this parameter
are "true" and "false".
.IP
[enum\-name\-1] [enum\-name\-2]...[enum\-name\-N]
.IP
Optional names for the enumerated values.
.IP
If the number of names listed is greater than
the "number\-of\-values" parameter, the "number\-of\-values"
will become the number of names.  If the number
of names is is less than the "number\-of\-values",
empty entries will be created.
.PP
\fB\-class\-create\-operation\fR
.PP
Create Operation Class header (.h) and implementation (.cxx) files.
.PP
Usage:  <operation\-class\-name>
.IP
<command\-line\-switch>
<short\-description>
[\-no\-parameters]
.IP
operation\-class\-name
.IP
Required name of the operation class that MUST start with "Operation"
.IP
command\-line\-switch
.IP
Required command line switch for operation.
.IP
short\-description
.IP
Required short description within double quotes.
.HP
\fB\-no\-parameters\fR
.IP
Optional parameter if the operation does not use parameters.
.PP
\fB\-convert\-affine\fR
CONVERT AN AFFINE FILE BETWEEN CONVENTIONS
.IP
wb_command \fB\-convert\-affine\fR
.IP
[\-from\-world] \- input is a NIFTI 'world' affine
.IP
<input> \- the input affine
.IP
[\-from\-flirt] \- input is a flirt matrix
.IP
<input> \- the input affine
<source\-volume> \- the source volume used when generating the input
.IP
affine
.IP
<target\-volume> \- the target volume used when generating the input
.IP
affine
.IP
[\-to\-world] \- write output as a NIFTI 'world' affine
.IP
<output> \- output \- the output affine
.IP
[\-to\-flirt] \- repeatable \- write output as a flirt matrix
.IP
<output> \- output \- the output affine
<source\-volume> \- the volume you want to apply the transform to
<target\-volume> \- the target space you want the transformed volume to
.IP
match
.IP
NIFTI world matrices can be used directly on mm coordinates via matrix
multiplication, they use the NIFTI coordinate system, that is, positive X
is right, positive Y is anterior, and positive Z is superior.
.IP
You must specify exactly one \fB\-from\fR option, but you may specify multiple
\fB\-to\fR options, and any \fB\-to\fR option that takes volumes may be specified more
than once.
.PP
\fB\-convert\-fiber\-orientations\fR
CONVERT BINGHAM PARAMETER VOLUMES TO FIBER ORIENTATION FILE
.IP
wb_command \fB\-convert\-fiber\-orientations\fR
.IP
<label\-volume> \- volume of cifti structure labels
<fiber\-out> \- output \- the output fiber orientation file
.IP
[\-fiber] \- repeatable \- specify the parameter volumes for a fiber
.IP
<mean\-f> \- mean fiber strength
<stdev\-f> \- standard deviation of fiber strength
<theta> \- theta angle
<phi> \- phi angle
<psi> \- psi angle
<ka> \- ka bingham parameter
<kb> \- kb bingham parameter
.IP
Takes precomputed bingham parameters from volume files and converts them
to the format workbench uses for display.  The <label\-volume> argument
must be a label volume, where the labels use these strings:
.IP
CORTEX_LEFT
CORTEX_RIGHT
CEREBELLUM
ACCUMBENS_LEFT
ACCUMBENS_RIGHT
ALL_GREY_MATTER
ALL_WHITE_MATTER
AMYGDALA_LEFT
AMYGDALA_RIGHT
BRAIN_STEM
CAUDATE_LEFT
CAUDATE_RIGHT
CEREBELLAR_WHITE_MATTER_LEFT
CEREBELLAR_WHITE_MATTER_RIGHT
CEREBELLUM_LEFT
CEREBELLUM_RIGHT
CEREBRAL_WHITE_MATTER_LEFT
CEREBRAL_WHITE_MATTER_RIGHT
CORTEX
DIENCEPHALON_VENTRAL_LEFT
DIENCEPHALON_VENTRAL_RIGHT
HIPPOCAMPUS_LEFT
HIPPOCAMPUS_RIGHT
INVALID
OTHER
OTHER_GREY_MATTER
OTHER_WHITE_MATTER
PALLIDUM_LEFT
PALLIDUM_RIGHT
PUTAMEN_LEFT
PUTAMEN_RIGHT
THALAMUS_LEFT
THALAMUS_RIGHT
.PP
\fB\-convert\-matrix4\-to\-matrix2\fR
GENERATES A MATRIX2 CIFTI FROM MATRIX4 WBSPARSE
.IP
wb_command \fB\-convert\-matrix4\-to\-matrix2\fR
.IP
<matrix4\-wbsparse> \- a wbsparse matrix4 file
<counts\-out> \- output \- the total fiber counts, as a cifti file
.IP
[\-distances] \- output average trajectory distance
.IP
<distance\-out> \- output \- the distances, as a cifti file
.IP
This command makes a cifti file from the fiber counts in a matrix4
wbsparse file, and optionally a second cifti file from the distances.
.PP
\fB\-convert\-matrix4\-to\-workbench\-sparse\fR
CONVERT A 3\-FILE MATRIX4 TO A WORKBENCH SPARSE FILE
.IP
wb_command \fB\-convert\-matrix4\-to\-workbench\-sparse\fR
.IP
<matrix4_1> \- the first matrix4 file
<matrix4_2> \- the second matrix4 file
<matrix4_3> \- the third matrix4 file
<orientation\-file> \- the .fiberTEMP.nii file this trajectory file applies
.IP
to
.IP
<voxel\-list> \- list of white matter voxel index triplets as used in the
.IP
trajectory matrix
.IP
<wb\-sparse\-out> \- output \- the output workbench sparse file
.IP
[\-surface\-seeds] \- specify the surface seed space
.IP
<seed\-roi> \- metric roi file of all vertices used in the seed space
.IP
[\-volume\-seeds] \- specify the volume seed space
.IP
<cifti\-template> \- cifti file to use the volume mappings from
<direction> \- dimension along the cifti file to take the mapping from,
.IP
ROW or COLUMN
.IP
Converts the matrix 4 output of probtrackx to workbench sparse file
format.  Exactly one of \fB\-surface\-seeds\fR and \fB\-volume\-seeds\fR must be
specified.
.PP
\fB\-convert\-warpfield\fR
CONVERT A WARPFIELD BETWEEN CONVENTIONS
.IP
wb_command \fB\-convert\-warpfield\fR
.IP
[\-from\-world] \- input is a NIFTI 'world' warpfield
.IP
<input> \- the input warpfield
.IP
[\-from\-fnirt] \- input is a fnirt warpfield
.IP
<input> \- the input warpfield
<source\-volume> \- the source volume used when generating the input
.IP
warpfield
.IP
[\-to\-world] \- write output as a NIFTI 'world' warpfield
.IP
<output> \- output \- the output warpfield
.IP
[\-to\-fnirt] \- repeatable \- write output as a fnirt warpfield
.IP
<output> \- output \- the output warpfield
<source\-volume> \- the volume you want to apply the warpfield to
.IP
NIFTI world warpfields can be used directly on mm coordinates via
sampling the three subvolumes at the coordinate and adding the sampled
values to the coordinate vector, they use the NIFTI coordinate system,
that is, X is left to right, Y is posterior to anterior, and Z is
inferior to superior.
.IP
NOTE: this command does not invert the warpfield, and to warp a surface,
you must use the inverse of the warpfield that warps the corresponding
volume.
.IP
You must specify exactly one \fB\-from\fR option, but you may specify multiple
\fB\-to\fR options, and \fB\-to\-fnirt\fR may be specified more than once.
.PP
\fB\-create\-signed\-distance\-volume\fR
CREATE SIGNED DISTANCE VOLUME FROM SURFACE
.IP
wb_command \fB\-create\-signed\-distance\-volume\fR
.IP
<surface> \- the input surface
<refspace> \- a volume in the desired output space (dims, spacing, origin)
<outvol> \- output \- the output volume
.IP
[\-roi\-out] \- output an roi volume of where the output has a computed
.IP
value
<roi\-vol> \- output \- the output roi volume
.IP
[\-fill\-value] \- specify a value to put in all voxels that don't get
.IP
assigned a distance
<value> \- value to fill with (default 0)
.IP
[\-exact\-limit] \- specify distance for exact output
.IP
<dist> \- distance in mm (default 5)
.IP
[\-approx\-limit] \- specify distance for approximate output
.IP
<dist> \- distance in mm (default 20)
.IP
[\-approx\-neighborhood] \- voxel neighborhood for approximate calculation
.IP
<num> \- size of neighborhood cube measured from center to face, in
.IP
voxels (default 2 = 5x5x5)
.IP
[\-winding] \- winding method for point inside surface test
.IP
<method> \- name of the method (default EVEN_ODD)
.TP
Computes the signed distance function of the surface.
Exact distance is
.IP
calculated by finding the closest point on any surface triangle to the
center of the voxel.  Approximate distance is calculated starting with
these distances, using dijkstra's method with a neighborhood of voxels.
Specifying too small of an exact distance may produce unexpected results.
Valid specifiers for winding methods are as follows:
.IP
EVEN_ODD (default)
NEGATIVE
NONZERO
NORMALS
.IP
The NORMALS method uses the normals of triangles and edges, or the
closest triangle hit by a ray from the point.  This method may be
slightly faster, but is only reliable for a closed surface that does not
cross through itself.  All other methods count entry (positive) and exit
(negative) crossings of a vertical ray from the point, then counts as
inside if the total is odd, negative, or nonzero, respectively.
.PP
\fB\-estimate\-fiber\-binghams\fR
ESTIMATE FIBER ORIENTATION DISTRIBUTIONS FROM BEDPOSTX SAMPLES
.IP
wb_command \fB\-estimate\-fiber\-binghams\fR
.IP
<merged_f1samples> \- fiber 1 strength samples
<merged_th1samples> \- fiber 1 theta samples
<merged_ph1samples> \- fiber 1 phi samples
<merged_f2samples> \- fiber 2 strength samples
<merged_th2samples> \- fiber 2 theta samples
<merged_ph2samples> \- fiber 2 phi samples
<merged_f3samples> \- fiber 3 strength samples
<merged_th3samples> \- fiber 3 theta samples
<merged_ph3samples> \- fiber 3 phi samples
<label\-volume> \- volume of cifti structure labels
<cifti\-out> \- output \- output cifti fiber distributons file
.IP
This command does an estimation of a bingham distribution for each fiber
orientation in each voxel which is labeled a structure identifier.  These
labelings come from the <label\-volume> argument, which must have labels
that match the following strings:
.IP
CORTEX_LEFT
CORTEX_RIGHT
CEREBELLUM
ACCUMBENS_LEFT
ACCUMBENS_RIGHT
ALL_GREY_MATTER
ALL_WHITE_MATTER
AMYGDALA_LEFT
AMYGDALA_RIGHT
BRAIN_STEM
CAUDATE_LEFT
CAUDATE_RIGHT
CEREBELLAR_WHITE_MATTER_LEFT
CEREBELLAR_WHITE_MATTER_RIGHT
CEREBELLUM_LEFT
CEREBELLUM_RIGHT
CEREBRAL_WHITE_MATTER_LEFT
CEREBRAL_WHITE_MATTER_RIGHT
CORTEX
DIENCEPHALON_VENTRAL_LEFT
DIENCEPHALON_VENTRAL_RIGHT
HIPPOCAMPUS_LEFT
HIPPOCAMPUS_RIGHT
INVALID
OTHER
OTHER_GREY_MATTER
OTHER_WHITE_MATTER
PALLIDUM_LEFT
PALLIDUM_RIGHT
PUTAMEN_LEFT
PUTAMEN_RIGHT
THALAMUS_LEFT
THALAMUS_RIGHT
.PP
\fB\-fiber\-dot\-products\fR
COMPUTE DOT PRODUCTS OF FIBER ORIENTATIONS WITH SURFACE NORMALS
.IP
wb_command \fB\-fiber\-dot\-products\fR
.IP
<white\-surf> \- the white/gray boundary surface
<fiber\-file> \- the fiber orientation file
<max\-dist> \- the maximum distance from any surface vertex a fiber
.IP
population may be, in mm
.IP
<direction> \- test against surface for whether a fiber population should
.IP
be used
.IP
<dot\-metric> \- output \- the metric of dot products
<f\-metric> \- output \- a metric of the f values of the fiber distributions
.IP
For each vertex, this command finds the closest fiber population that
satisfies the <direction> test, and computes the absolute value of the
dot product of the surface normal and the normalized mean direction of
each fiber.  The <direction> test must be one of INSIDE, OUTSIDE, or ANY,
which causes the command to only use fiber populations that are inside
the surface, outside the surface, or to not care which direction it is
from the surface.  Each fiber population is output in a separate metric
column.
.PP
\fB\-file\-convert\fR
CHANGE VERSION OF FILE FORMAT
.IP
wb_command \fB\-file\-convert\fR
.IP
[\-border\-version\-convert] \- write a border file with a different version
.IP
<border\-in> \- the input border file
<out\-version> \- the format version to write as, 1 or 3 (2 doesn't
.IP
exist)
.IP
<border\-out> \- output \- the output border file
.IP
[\-surface] \- must be specified if the input is version 1
.IP
<surface> \- use this surface file for structure and number of
.IP
vertices, ignore borders on other structures
.IP
[\-nifti\-version\-convert] \- write a nifti file with a different version
.IP
<input> \- the input nifti file
<version> \- the nifti version to write as
<output> \- output \- the output nifti file
.IP
[\-cifti\-version\-convert] \- write a cifti file with a different version
.IP
<cifti\-in> \- the input cifti file
<version> \- the cifti version to write as
<cifti\-out> \- output \- the output cifti file
.IP
You may only specify one top\-level option.
.PP
\fB\-file\-information\fR
LIST INFORMATION ABOUT A FILE'S CONTENT
.IP
wb_command \fB\-file\-information\fR
.IP
<data\-file> \- data file
.IP
[\-no\-map\-info] \- do not show map information for files that support maps
.IP
[\-only\-step\-interval] \- suppress normal output, print the interval
.IP
between maps
.IP
[\-only\-number\-of\-maps] \- suppress normal output, print the number of maps
.IP
[\-only\-map\-names] \- suppress normal output, print the names of all maps
.IP
[\-only\-metadata] \- suppress normal output, print file metadata
.IP
[\-key] \- only print the metadata for one key, with no formatting
.IP
<key> \- the metadata key
.TP
List information about the content of a data file.
Only one \fB\-only\fR option
.TP
may be specified.
The information listed when no \fB\-only\fR option is present
.IP
is dependent upon the type of data file.
.PP
\fB\-foci\-get\-projection\-vertex\fR
GET PROJECTION VERTEX FOR FOCI
.IP
wb_command \fB\-foci\-get\-projection\-vertex\fR
.IP
<foci> \- the foci file
<surface> \- the surface related to the foci file
<metric\-out> \- output \- the output metric file
.IP
[\-name] \- select a focus by name
.IP
<name> \- the name of the focus
.IP
For each focus, a column is created in <metric\-out>, and the vertex with
the most influence on its projection is assigned a value of 1 in that
column, with all other vertices 0.  If \fB\-name\fR is used, only one focus will
be used.
.PP
\fB\-foci\-list\-coords\fR
OUTPUT FOCI COORDINATES IN A TEXT FILE
.IP
wb_command \fB\-foci\-list\-coords\fR
.IP
<foci\-file> \- input foci file
<coord\-file\-out> \- output \- the output coordinate text file
.IP
[\-names\-out] \- output the foci names
.IP
<names\-file\-out> \- output \- text file to put foci names in
.IP
Output the coordinates for every focus in the foci file, and optionally
the focus names in a second text file.
.PP
\fB\-foci\-resample\fR
PROJECT FOCI TO A DIFFERENT SURFACE
.IP
wb_command \fB\-foci\-resample\fR
.IP
<foci\-in> \- the input foci file
<foci\-out> \- output \- the output foci file
.IP
[\-left\-surfaces] \- the left surfaces for resampling
.IP
<current\-surf> \- the surface the foci are currently projected on
<new\-surf> \- the surface to project the foci onto
.IP
[\-right\-surfaces] \- the right surfaces for resampling
.IP
<current\-surf> \- the surface the foci are currently projected on
<new\-surf> \- the surface to project the foci onto
.IP
[\-cerebellum\-surfaces] \- the cerebellum surfaces for resampling
.IP
<current\-surf> \- the surface the foci are currently projected on
<new\-surf> \- the surface to project the foci onto
.IP
[\-discard\-distance\-from\-surface] \- ignore the distance the foci are above
.IP
or below the current surface
.IP
[\-restore\-xyz] \- put the original xyz coordinates into the foci, rather
.IP
than the coordinates obtained from unprojection
.IP
Unprojects foci from the <current\-surf> for the structure, then projects
them to <new\-surf>.  If the foci have meaningful distances above or below
the surface, use anatomical surfaces.  If the foci should be on the
surface, use registered spheres and the options
\fB\-discard\-distance\-from\-surface\fR and \fB\-restore\-xyz\fR.
.PP
\fB\-gifti\-all\-labels\-to\-rois\fR
MAKE ROIS FROM ALL LABELS IN A GIFTI COLUMN
.IP
wb_command \fB\-gifti\-all\-labels\-to\-rois\fR
.IP
<label\-in> \- the input gifti label file
<map> \- the number or name of the label map to use
<metric\-out> \- output \- the output metric file
.IP
The output metric file has a column for each label in the specified input
map, other than the ??? label, each of which contains an ROI of all
vertices that are set to the corresponding label.
.PP
\fB\-gifti\-convert\fR
CONVERT A GIFTI FILE TO A DIFFERENT ENCODING
.IP
wb_command \fB\-gifti\-convert\fR
.IP
<gifti\-encoding> \- what the output encoding should be
<input\-gifti\-file> \- the input gifti file
<output\-gifti\-file> \- output \- the output gifti file
.IP
The value of <gifti\-encoding> must be one of the following:
.IP
ASCII
BASE64_BINARY
GZIP_BASE64_BINARY
EXTERNAL_FILE_BINARY
.PP
\fB\-gifti\-label\-add\-prefix\fR
ADD PREFIX TO ALL LABEL NAMES IN A GIFTI LABEL FILE
.IP
wb_command \fB\-gifti\-label\-add\-prefix\fR
.IP
<label\-in> \- the input label file
<prefix> \- the prefix string to add
<label\-out> \- output \- the output label file
.IP
For each label other than '???', prepend <prefix> to the label name.
.PP
\fB\-gifti\-label\-to\-roi\fR
MAKE A GIFTI LABEL INTO AN ROI METRIC
.IP
wb_command \fB\-gifti\-label\-to\-roi\fR
.IP
<label\-in> \- the input gifti label file
<metric\-out> \- output \- the output metric file
.IP
[\-name] \- select label by name
.IP
<label\-name> \- the label name that you want an roi of
.IP
[\-key] \- select label by key
.IP
<label\-key> \- the label key that you want an roi of
.IP
[\-map] \- select a single label map to use
.IP
<map> \- the map number or name
.IP
For each map in <label\-in>, a map is created in <metric\-out> where all
locations labeled with <label\-name> or with a key of <label\-key> are
given a value of 1, and all other locations are given 0.  Exactly one of
\fB\-name\fR and \fB\-key\fR must be specified.  Specify \fB\-map\fR to use only one map from
<label\-in>.
.PP
\fB\-label\-dilate\fR
DILATE A LABEL FILE
.IP
wb_command \fB\-label\-dilate\fR
.IP
<label> \- the input label
<surface> \- the surface to dilate on
<dilate\-dist> \- distance in mm to dilate the labels
<label\-out> \- output \- the output label file
.IP
[\-bad\-vertex\-roi] \- specify an roi of vertices to overwrite, rather than
.IP
vertices with the unlabeled key
<roi\-metric> \- metric file, positive values denote vertices to have
.IP
their values replaced
.IP
[\-column] \- select a single column to dilate
.IP
<column> \- the column number or name
.IP
[\-corrected\-areas] \- vertex areas to use instead of computing them from
.IP
the surface
<area\-metric> \- the corrected vertex areas, as a metric
.IP
Fills in label information for all vertices designated as bad, up to the
specified distance away from other labels.  If \fB\-bad\-vertex\-roi\fR is
specified, all vertices, including those with the unlabeled key, are
good, except for vertices with a positive value in the ROI.  If it is not
specified, only vertices with the unlabeled key are bad.
.PP
\fB\-label\-erode\fR
ERODE A LABEL FILE
.IP
wb_command \fB\-label\-erode\fR
.IP
<label> \- the input label
<surface> \- the surface to erode on
<erode\-dist> \- distance in mm to erode the labels
<label\-out> \- output \- the output label file
.IP
[\-roi] \- assume values outside this roi are labeled
.IP
<roi\-metric> \- metric file, positive values denote vertices that have
.IP
data
.IP
[\-column] \- select a single column to erode
.IP
<column> \- the column number or name
.IP
[\-corrected\-areas] \- vertex areas to use instead of computing them from
.IP
the surface
<area\-metric> \- the corrected vertex areas, as a metric
.IP
Around each vertex that is unlabeled, set surrounding vertices to
unlabeled.  The surrounding vertices are all immediate neighbors and all
vertices within the specified distance.
.IP
Note that the \fB\-corrected\-areas\fR option uses an approximate correction for
distance along the surface.
.PP
\fB\-label\-export\-table\fR
EXPORT LABEL TABLE FROM GIFTI AS TEXT
.IP
wb_command \fB\-label\-export\-table\fR
.IP
<label\-in> \- the input label file
<table\-out> \- output \- the output text file
.IP
Takes the label table from the gifti label file, and writes it to a text
format matching what is expected by \fB\-metric\-label\-import\fR.
.PP
\fB\-label\-mask\fR
MASK A LABEL FILE
.IP
wb_command \fB\-label\-mask\fR
.IP
<label> \- the label file to mask
<mask> \- the mask metric
<label\-out> \- output \- the output label file
.IP
[\-column] \- select a single column
.IP
<column> \- the column number or name
.IP
By default, the output label is a copy of the input label, but with the
\&'unused' label wherever the mask metric is not positive.  if \fB\-column\fR is
specified, the output contains only one column, the masked version of the
specified input column.
.PP
\fB\-label\-merge\fR
MERGE LABEL FILES INTO A NEW FILE
.IP
wb_command \fB\-label\-merge\fR
.IP
<label\-out> \- output \- the output label
.IP
[\-label] \- repeatable \- specify an input label
.IP
<label\-in> \- a label file to use columns from
.IP
[\-column] \- repeatable \- select a single column to use
.IP
<column> \- the column number or name
.IP
[\-up\-to] \- use an inclusive range of columns
.IP
<last\-column> \- the number or name of the last column to include
.IP
[\-reverse] \- use the range in reverse order
.IP
Takes one or more label files and constructs a new label file by
concatenating columns from them.  The input files must have the same
number of vertices and the same structure.
.IP
Example: wb_command \fB\-label\-merge\fR out.label.gii \fB\-label\fR first.label.gii
\fB\-column\fR 1 \fB\-label\fR second.label.gii
.IP
This example would take the first column from first.label.gii and all
subvolumes from second.label.gii, and write these to out.label.gii.
.PP
\fB\-label\-modify\-keys\fR
CHANGE KEY VALUES IN A LABEL FILE
.IP
wb_command \fB\-label\-modify\-keys\fR
.IP
<label\-in> \- the input label file
<remap\-file> \- text file with old and new key values
<label\-out> \- output \- output label file
.IP
[\-column] \- select a single column to use
.IP
<column> \- the column number or name
.IP
<remap\-file> should have lines of the form 'oldkey newkey', like so:
.IP
3 5
5 8
8 2
.IP
This would change the current label with key '3' to use the key '5'
instead, 5 would use 8, and 8 would use 2.  Any collision in key values
results in the label that was not specified in the remap file getting
remapped to an otherwise unused key.  Remapping more than one key to the
same new key, or the same key to more than one new key, results in an
error.  This will not change the appearance of the file when displayed,
it will change the keys in the data at the same time.
.PP
\fB\-label\-resample\fR
RESAMPLE A LABEL FILE TO A DIFFERENT MESH
.IP
wb_command \fB\-label\-resample\fR
.IP
<label\-in> \- the label file to resample
<current\-sphere> \- a sphere surface with the mesh that the label file is
.IP
currently on
.IP
<new\-sphere> \- a sphere surface that is in register with <current\-sphere>
.IP
and has the desired output mesh
.IP
<method> \- the method name
<label\-out> \- output \- the output label file
.IP
[\-area\-surfs] \- specify surfaces to do vertex area correction based on
.IP
<current\-area> \- a relevant anatomical surface with <current\-sphere>
.IP
mesh
.IP
<new\-area> \- a relevant anatomical surface with <new\-sphere> mesh
.IP
[\-area\-metrics] \- specify vertex area metrics to do area correction based
.IP
on
<current\-area> \- a metric file with vertex areas for <current\-sphere>
.IP
mesh
.IP
<new\-area> \- a metric file with vertex areas for <new\-sphere> mesh
.IP
[\-current\-roi] \- use an input roi on the current mesh to exclude non\-data
.IP
vertices
<roi\-metric> \- the roi, as a metric file
.IP
[\-valid\-roi\-out] \- output the ROI of vertices that got data from valid
.IP
source vertices
<roi\-out> \- output \- the output roi as a metric
.IP
[\-largest] \- use only the label of the vertex with the largest weight
.IP
Resamples a label file, given two spherical surfaces that are in
register.  If ADAP_BARY_AREA is used, exactly one of \fB\-area\-surfs\fR or
\fB\-area\-metrics\fR must be specified.
.IP
The ADAP_BARY_AREA method is recommended for label data, because it
should be better at resolving vertices that are near multiple labels, or
in case of downsampling.  Midthickness surfaces are recommended for the
vertex areas for most data.
.IP
The \fB\-largest\fR option results in nearest vertex behavior when used with
BARYCENTRIC, as it uses the value of the source vertex that has the
largest weight.
.IP
When \fB\-largest\fR is not specified, the vertex weights are summed according
to which label they correspond to, and the label with the largest sum is
used.
.IP
The <method> argument must be one of the following:
.IP
ADAP_BARY_AREA
BARYCENTRIC
.PP
\fB\-label\-to\-border\fR
DRAW BORDERS AROUND LABELS
.IP
wb_command \fB\-label\-to\-border\fR
.IP
<surface> \- the surface to use for neighbor information
<label\-in> \- the input label file
<border\-out> \- output \- the output border file
.IP
[\-placement] \- set how far along the edge border points are drawn
.IP
<fraction> \- fraction along edge from inside vertex (default 0.33)
.IP
[\-column] \- select a single column
.IP
<column> \- the column number or name
.IP
For each label, finds all edges on the mesh that cross the boundary of
the label, and draws borders through them.  By default, this is done on
all columns in the input file, using the map name as the class name for
the border.
.PP
\fB\-label\-to\-volume\-mapping\fR
MAP LABEL FILE TO VOLUME
.IP
wb_command \fB\-label\-to\-volume\-mapping\fR
.IP
<label> \- the input label file
<surface> \- the surface to use coordinates from
<volume\-space> \- a volume file in the desired output volume space
<volume\-out> \- output \- the output volume file
.IP
[\-nearest\-vertex] \- use the label from the vertex closest to the voxel
.IP
center
<distance> \- how far from the surface to map labels to voxels, in mm
.IP
[\-ribbon\-constrained] \- use ribbon constrained mapping algorithm
.IP
<inner\-surf> \- the inner surface of the ribbon
<outer\-surf> \- the outer surface of the ribbon
.IP
[\-voxel\-subdiv] \- voxel divisions while estimating voxel weights
.IP
<subdiv\-num> \- number of subdivisions, default 3
.TP
Maps labels from a gifti label file into a volume file.
You must specify
.TP
exactly one mapping method option.
The \fB\-nearest\-vertex\fR method uses the
.TP
label from the vertex closest to the voxel center.
The
.HP
\fB\-ribbon\-constrained\fR method uses the same method as in
.HP
\fB\-volume\-to\-surface\-mapping\fR, then uses the weights in reverse, with
.IP
popularity logic to decide on a label to use.
.PP
\fB\-metadata\-remove\-provenance\fR
REMOVE PROVENANCE INFORMATION FROM FILE METADATA
.IP
wb_command \fB\-metadata\-remove\-provenance\fR
.IP
<input\-file> \- the file to remove provenance information from
<output\-file> \- output \- the name to save the modified file as
.IP
Removes the provenance metadata fields added by workbench during
processing.
.PP
\fB\-metadata\-string\-replace\fR
REPLACE A STRING IN ALL METADATA OF A FILE
.IP
wb_command \fB\-metadata\-string\-replace\fR
.IP
<input\-file> \- the file to replace metadata in
<find\-string> \- the string to find
<replace\-string> \- the string to replace <find\-string> with
<output\-file> \- output \- the name to save the modified file as
.IP
[\-case\-insensitive] \- match with case variation also
.IP
Replaces all occurrences of <find\-string> in the metadata and map names
of <input\-file> with <replace\-string>.
.PP
\fB\-metric\-convert\fR
CONVERT METRIC FILE TO FAKE NIFTI
.IP
wb_command \fB\-metric\-convert\fR
.IP
[\-to\-nifti] \- convert metric to nifti
.IP
<metric\-in> \- the metric to convert
<nifti\-out> \- output \- the output nifti file
.IP
[\-from\-nifti] \- convert nifti to metric
.IP
<nifti\-in> \- the nifti file to convert
<surface\-in> \- surface file to use number of vertices and structure
.IP
from
.IP
<metric\-out> \- output \- the output metric file
.IP
The purpose of this command is to convert between metric files and nifti1
so that gifti\-unaware programs can operate on the data.  You must specify
exactly one of the options.
.PP
\fB\-metric\-dilate\fR
DILATE A METRIC FILE
.IP
wb_command \fB\-metric\-dilate\fR
.IP
<metric> \- the metric to dilate
<surface> \- the surface to compute on
<distance> \- distance in mm to dilate
<metric\-out> \- output \- the output metric
.IP
[\-bad\-vertex\-roi] \- specify an roi of vertices to overwrite, rather than
.IP
vertices with value zero
<roi\-metric> \- metric file, positive values denote vertices to have
.IP
their values replaced
.IP
[\-data\-roi] \- specify an roi of where there is data
.IP
<roi\-metric> \- metric file, positive values denote vertices that have
.IP
data
.IP
[\-column] \- select a single column to dilate
.IP
<column> \- the column number or name
.IP
[\-nearest] \- use the nearest good value instead of a weighted average
.IP
[\-linear] \- fill in values with linear interpolation along strongest
.IP
gradient
.IP
[\-exponent] \- use a different exponent in the weighting function
.IP
<exponent> \- exponent 'n' to use in (area / (distance ^ n)) as the
.IP
weighting function (default 2)
.IP
[\-corrected\-areas] \- vertex areas to use instead of computing them from
.IP
the surface
<area\-metric> \- the corrected vertex areas, as a metric
.IP
For all metric vertices that are designated as bad, if they neighbor a
non\-bad vertex with data or are within the specified distance of such a
vertex, replace the value with a distance weighted average of nearby
non\-bad vertices that have data, otherwise set the value to zero.  No
matter how small <distance> is, dilation will always use at least the
immediate neighbor vertices.  If \fB\-nearest\fR is specified, it will use the
value from the closest non\-bad vertex with data within range instead of a
weighted average.
.IP
If \fB\-bad\-vertex\-roi\fR is specified, only vertices with a positive value in
the ROI are bad.  If it is not specified, only vertices that have data,
with a value of zero, are bad.  If \fB\-data\-roi\fR is not specified, all
vertices are assumed to have data.
.IP
Note that the \fB\-corrected\-areas\fR option uses an approximate correction for
the change in distances along a group average surface.
.PP
\fB\-metric\-erode\fR
ERODE A METRIC FILE
.IP
wb_command \fB\-metric\-erode\fR
.IP
<metric> \- the metric file to erode
<surface> \- the surface to compute on
<distance> \- distance in mm to erode
<metric\-out> \- output \- the output metric
.IP
[\-roi] \- assume values outside this roi are nonzero
.IP
<roi\-metric> \- metric file, positive values denote vertices that have
.IP
data
.IP
[\-column] \- select a single column to erode
.IP
<column> \- the column number or name
.IP
[\-corrected\-areas] \- vertex areas to use instead of computing them from
.IP
the surface
<area\-metric> \- the corrected vertex areas, as a metric
.IP
Around each vertex with a value of zero, set surrounding vertices to
zero.  The surrounding vertices are all immediate neighbors and all
vertices within the specified distance.
.IP
Note that the \fB\-corrected\-areas\fR option uses an approximate correction for
distance along the surface.
.PP
\fB\-metric\-estimate\-fwhm\fR
ESTIMATE FWHM SMOOTHNESS OF A METRIC FILE
.IP
wb_command \fB\-metric\-estimate\-fwhm\fR
.IP
<surface> \- the surface to use for distance and neighbor information
<metric\-in> \- the input metric
.IP
[\-roi] \- use only data within an ROI
.IP
<roi\-metric> \- the metric file to use as an ROI
.IP
[\-column] \- select a single column to estimate smoothness of
.IP
<column> \- the column number or name
.IP
Estimates the smoothness of the metric columns, printing the estimates to
standard output.  These estimates ignore variation in vertex spacing.
.PP
\fB\-metric\-extrema\fR
FIND EXTREMA IN A METRIC FILE
.IP
wb_command \fB\-metric\-extrema\fR
.IP
<surface> \- the surface to use for distance information
<metric\-in> \- the metric to find the extrema of
<distance> \- the minimum distance between identified extrema of the same
.IP
type
.IP
<metric\-out> \- output \- the output extrema metric
.IP
[\-presmooth] \- smooth the metric before finding extrema
.IP
<kernel> \- the sigma for the gaussian smoothing kernel, in mm
.IP
[\-roi] \- ignore values outside the selected area
.IP
<roi\-metric> \- the area to find extrema in, as a metric
.IP
[\-threshold] \- ignore small extrema
.IP
<low> \- the largest value to consider for being a minimum
<high> \- the smallest value to consider for being a maximum
.IP
[\-sum\-columns] \- output the sum of the extrema columns instead of each
.IP
column separately
.IP
[\-consolidate\-mode] \- use consolidation of local minima instead of a
.IP
large neighborhood
.IP
[\-only\-maxima] \- only find the maxima
.IP
[\-only\-minima] \- only find the minima
.IP
[\-column] \- select a single column to find extrema in
.IP
<column> \- the column number or name
.IP
Finds extrema in a metric file, such that no two extrema of the same type
are within <distance> of each other.  The extrema are labeled as \fB\-1\fR for
minima, 1 for maxima, 0 otherwise.  If \fB\-only\-maxima\fR or \fB\-only\-minima\fR is
specified, then it will ignore extrema not of the specified type.  These
options are mutually exclusive.
.IP
If \fB\-roi\fR is specified, not only is data outside the roi not used, but any
vertex on the edge of the ROI will never be counted as an extrema, in
case the ROI cuts across a gradient, which would otherwise generate
extrema where there should be none.
.IP
If \fB\-sum\-columns\fR is specified, these extrema columns are summed, and the
output has a single column with this result.
.IP
By default, a datapoint is an extrema only if it is more extreme than
every other datapoint that is within <distance> from it.  If
\fB\-consolidate\-mode\fR is used, it instead starts by finding all datapoints
that are more extreme than their immediate neighbors, then while there
are any extrema within <distance> of each other, take the two extrema
closest to each other and merge them into one by a weighted average based
on how many original extrema have been merged into each.
.IP
By default, all input columns are used with no smoothing, use \fB\-column\fR to
specify a single column to use, and \fB\-presmooth\fR to smooth the input before
finding the extrema.
.PP
\fB\-metric\-false\-correlation\fR
COMPARE CORRELATION LOCALLY AND ACROSS/THROUGH SULCI/GYRI
.IP
wb_command \fB\-metric\-false\-correlation\fR
.IP
<surface> \- the surface to compute geodesic and 3D distance with
<metric\-in> \- the metric to correlate
<3D\-dist> \- maximum 3D distance to check around each vertex
<geo\-outer> \- maximum geodesic distance to use for neighboring
.IP
correlation
.IP
<geo\-inner> \- minimum geodesic distance to use for neighboring
.IP
correlation
.IP
<metric\-out> \- output \- the output metric
.IP
[\-roi] \- select a region of interest that has data
.IP
<roi\-metric> \- the region, as a metric file
.IP
[\-dump\-text] \- dump the raw measures used to a text file
.IP
<text\-out> \- the output text file
.IP
For each vertex, compute the average correlation within a range of
geodesic distances that don't cross a sulcus/gyrus, and the correlation
to the closest vertex crossing a sulcus/gyrus.  A vertex is considered to
cross a sulcus/gyrus if the 3D distance is less than a third of the
geodesic distance.  The output file contains the ratio between these
correlations, and some additional maps to help explain the ratio.
.PP
\fB\-metric\-fill\-holes\fR
FILL HOLES IN AN ROI METRIC
.IP
wb_command \fB\-metric\-fill\-holes\fR
.IP
<surface> \- the surface to use for neighbor information
<metric\-in> \- the input ROI metric
<metric\-out> \- output \- the output ROI metric
.IP
[\-corrected\-areas] \- vertex areas to use instead of computing them from
.IP
the surface
<area\-metric> \- the corrected vertex areas, as a metric
.IP
Finds all connected areas that are not included in the ROI, and writes
ones into all but the largest one, in terms of surface area.
.PP
\fB\-metric\-find\-clusters\fR
FILTER CLUSTERS BY SURFACE AREA
.IP
wb_command \fB\-metric\-find\-clusters\fR
.IP
<surface> \- the surface to compute on
<metric\-in> \- the input metric
<value\-threshold> \- threshold for data values
<minimum\-area> \- threshold for cluster area, in mm^2
<metric\-out> \- output \- the output metric
.IP
[\-less\-than] \- find values less than <value\-threshold>, rather than
.IP
greater
.IP
[\-roi] \- select a region of interest
.IP
<roi\-metric> \- the roi, as a metric
.IP
[\-corrected\-areas] \- vertex areas to use instead of computing them from
.IP
the surface
<area\-metric> \- the corrected vertex areas, as a metric
.IP
[\-column] \- select a single column
.IP
<column> \- the column number or name
.IP
[\-size\-ratio] \- ignore clusters smaller than a given fraction of the
.IP
largest cluster in map
<ratio> \- fraction of the largest cluster's area
.IP
[\-distance] \- ignore clusters further than a given distance from the
.IP
largest cluster
<distance> \- how far from the largest cluster a cluster can be, edge
.IP
to edge, in mm
.IP
[\-start] \- start labeling clusters from a value other than 1
.IP
<startval> \- the value to give the first cluster found
.IP
Outputs a metric with nonzero integers for all vertices within a large
enough cluster, and zeros elsewhere.  The integers denote cluster
membership (by default, first cluster found will use value 1, second
cluster 2, etc).  By default, values greater than <value\-threshold> are
considered to be in a cluster, use \fB\-less\-than\fR to test for values less
than the threshold.  To apply this as a mask to the data, or to do more
complicated thresholding, see \fB\-metric\-math\fR.
.PP
\fB\-metric\-gradient\fR
SURFACE GRADIENT OF A METRIC FILE
.IP
wb_command \fB\-metric\-gradient\fR
.IP
<surface> \- the surface to compute the gradient on
<metric\-in> \- the metric to compute the gradient of
<metric\-out> \- output \- the magnitude of the gradient
.IP
[\-presmooth] \- smooth the metric before computing the gradient
.IP
<kernel> \- the sigma for the gaussian smoothing kernel, in mm
.IP
[\-roi] \- select a region of interest to take the gradient of
.IP
<roi\-metric> \- the area to take the gradient within, as a metric
.IP
[\-match\-columns] \- for each input column, use the corresponding column
.IP
from the roi
.IP
[\-vectors] \- output gradient vectors
.IP
<vector\-metric\-out> \- output \- the vectors as a metric file
.IP
[\-column] \- select a single column to compute the gradient of
.IP
<column> \- the column number or name
.IP
[\-corrected\-areas] \- vertex areas to use instead of computing them from
.IP
the surface
<area\-metric> \- the corrected vertex areas, as a metric
.IP
[\-average\-normals] \- average the normals of each vertex with its
.IP
neighbors before using them to compute the gradient
.IP
At each vertex, the immediate neighbors are unfolded onto a plane tangent
to the surface at the vertex (specifically, perpendicular to the normal).
The gradient is computed using a regression between the unfolded
positions of the vertices and their values.  The gradient is then given
by the slopes of the regression, and reconstructed as a 3D gradient
vector.  By default, takes the gradient of all columns, with no
presmoothing, across the whole surface, without averaging the normals of
the surface among neighbors.
.IP
When using \fB\-corrected\-areas\fR, note that it is an approximate correction.
Doing smoothing on individual surfaces before averaging/gradient is
preferred, when possible, in order to make use of the original surface
structure.
.IP
Specifying an ROI will restrict the gradient to only use data from where
the ROI metric is positive, and output zeros anywhere the ROI metric is
not positive.
.IP
By default, the first column of the roi metric is used for all input
columns.  When \fB\-match\-columns\fR is specified to the \fB\-roi\fR option, the input
and roi metrics must have the same number of columns, and for each input
column's index, the same column index is used in the roi metric.  If the
\fB\-match\-columns\fR option to \fB\-roi\fR is used while the \fB\-column\fR option is also
used, the number of columns of the roi metric must match the input
metric, and it will use the roi column with the index of the selected
input column.
.IP
The vector output metric is organized such that the X, Y, and Z
components from a single input column are consecutive columns.
.PP
\fB\-metric\-label\-import\fR
IMPORT A GIFTI LABEL FILE FROM A METRIC FILE
.IP
wb_command \fB\-metric\-label\-import\fR
.IP
<input> \- the input metric file
<label\-list\-file> \- text file containing the values and names for labels
<output> \- output \- the output gifti label file
.IP
[\-discard\-others] \- set any values not mentioned in the label list to the
.IP
??? label
.IP
[\-unlabeled\-value] \- set the value that will be interpreted as unlabeled
.IP
<value> \- the numeric value for unlabeled (default 0)
.IP
[\-column] \- select a single column to import
.IP
<column> \- the column number or name
.IP
[\-drop\-unused\-labels] \- remove any unused label values from the label
.IP
table
.IP
Creates a new gifti label file from a metric file with label\-like values.
You may specify the empty string ('' will work on linux/mac) for
<label\-list\-file>, which will be treated as if it is an empty file.  The
label list file must have lines of the following format:
.IP
<labelname>
<value> <red> <green> <blue> <alpha>
.IP
Do not specify the "unlabeled" key in the file, it is assumed that 0
means not labeled unless \fB\-unlabeled\-value\fR is specified.  Label names must
be on a separate line, but may contain spaces or other unusual characters
(but not newline).  Whitespace is trimmed from both ends of the label
name, but is kept if it is in the middle of a label.  The values of red,
green, blue and alpha must be integers from 0 to 255, and will specify
the color the label is drawn as (alpha of 255 means opaque, which is
probably what you want).  By default, it will set new label names with
names of LABEL_# for any values encountered that are not mentioned in the
list file, specify \fB\-discard\-others\fR to instead set these voxels to the
"unlabeled" key.
.PP
\fB\-metric\-mask\fR
MASK A METRIC FILE
.IP
wb_command \fB\-metric\-mask\fR
.IP
<metric> \- the input metric
<mask> \- the mask metric
<metric\-out> \- output \- the output metric
.IP
[\-column] \- select a single column
.IP
<column> \- the column number or name
.IP
By default, the output metric is a copy of the input metric, but with
zeros wherever the mask metric is not positive.  if \fB\-column\fR is specified,
the output contains only one column, the masked version of the specified
input column.
.PP
\fB\-metric\-math\fR
EVALUATE EXPRESSION ON METRIC FILES
.IP
wb_command \fB\-metric\-math\fR
.IP
<expression> \- the expression to evaluate, in quotes
<metric\-out> \- output \- the output metric
.IP
[\-fixnan] \- replace NaN results with a value
.IP
<replace> \- value to replace NaN with
.IP
[\-var] \- repeatable \- a metric to use as a variable
.IP
<name> \- the name of the variable, as used in the expression
<metric> \- the metric file to use as this variable
.IP
[\-column] \- select a single column
.IP
<column> \- the column number or name
.IP
[\-repeat] \- reuse a single column for each column of calculation
.IP
This command evaluates <expression> at each surface vertex independently.
There must be at least one \fB\-var\fR option (to get the structure, number of
vertices, and number of columns from), even if the <name> specified in it
isn't used in <expression>.  All metrics must have the same number of
vertices.  Filenames are not valid in <expression>, use a variable name
and a \fB\-var\fR option with matching <name> to specify an input file.  If the
\fB\-column\fR option is given to any \fB\-var\fR option, only one column is used from
that file.  If \fB\-repeat\fR is specified, the file must either have only one
column, or have the \fB\-column\fR option specified.  All files that don't use
\fB\-repeat\fR must have the same number of columns requested to be used.  The
format of <expression> is as follows:
.IP
Expressions consist of constants, variables, operators, parentheses, and
functions, in infix notation, such as 'exp(\fB\-x\fR + 3) * scale'.  Variables
are strings of any length, using the characters a\-z, A\-Z, 0\-9, and _, but
may not take the name of a named constant.  Currently, there is only one
named constant, PI.  The operators are +, \-, *, /, ^, >, <, >=, <=, ==,
!=, !, &&, ||.  These behave as in C, except that ^ is exponentiation,
i.e. pow(x, y), and takes higher precedence than other binary operators
(also, '\-3^\-4^\-5' means '\-(3^(\-(4^\-5)))').  The <=, >=, ==, and !=
operators are given a small amount of wiggle room, equal to one millionth
of the smaller of the absolute values of the values being compared.
.IP
Comparison and logical operators return 0 or 1, you can do masking with
expressions like 'x * (mask > 0)'.  For all logical operators, an input
is considered true iff it is greater than 0.  The expression '0 < x < 5'
is not syntactically wrong, but it will NOT do what is desired, because
it is evaluated left to right, i.e. '((0 < x) < 5)', which will always
return 1, as both possible results of a comparison are less than 5.  A
warning is generated if an expression of this type is detected.  Use
something like 'x > 0 && x < 5' to get the desired behavior.
.IP
Whitespace between elements is ignored, ' sin ( 2 * x ) ' is equivalent
to 'sin(2*x)', but 's in(2*x)' is an error.  Implied multiplication is
not allowed, the expression '2x' will be parsed as a variable.
Parentheses are (), do not use [] or {}.  Functions require parentheses,
the expression 'sin x' is an error.
.IP
The following functions are supported:
.IP
sin: 1 argument, the sine of the argument (units are radians)
cos: 1 argument, the cosine of the argument (units are radians)
tan: 1 argument, the tangent of the argument (units are radians)
asin: 1 argument, the inverse of sine of the argument, in radians
acos: 1 argument, the inverse of cosine of the argument, in radians
atan: 1 argument, the inverse of tangent of the argument, in radians
atan2: 2 arguments, atan2(y, x) returns the inverse of tangent of
.IP
(y/x), in radians, determining quadrant by the sign of both arguments
.IP
sinh: 1 argument, the hyperbolic sine of the argument
cosh: 1 argument, the hyperbolic cosine of the argument
tanh: 1 argument, the hyperboloc tangent of the argument
asinh: 1 argument, the inverse hyperbolic sine of the argument
acosh: 1 argument, the inverse hyperbolic cosine of the argument
atanh: 1 argument, the inverse hyperboloc tangent of the argument
ln: 1 argument, the natural logarithm of the argument
exp: 1 argument, the constant e raised to the power of the argument
log: 1 argument, the base 10 logarithm of the argument
sqrt: 1 argument, the square root of the argument
abs: 1 argument, the absolute value of the argument
floor: 1 argument, the largest integer not greater than the argument
round: 1 argument, the nearest integer, with ties rounded away from
.IP
zero
.IP
ceil: 1 argument, the smallest integer not less than the argument
min: 2 arguments, min(x, y) returns y if (x > y), x otherwise
max: 2 arguments, max(x, y) returns y if (x < y), x otherwise
mod: 2 arguments, mod(x, y) = x \- y * floor(x / y), or 0 if y == 0
clamp: 3 arguments, clamp(x, low, high) = min(max(x, low), high)
.PP
\fB\-metric\-merge\fR
MERGE METRIC FILES INTO A NEW FILE
.IP
wb_command \fB\-metric\-merge\fR
.IP
<metric\-out> \- output \- the output metric
.IP
[\-metric] \- repeatable \- specify an input metric
.IP
<metric\-in> \- a metric file to use columns from
.IP
[\-column] \- repeatable \- select a single column to use
.IP
<column> \- the column number or name
.IP
[\-up\-to] \- use an inclusive range of columns
.IP
<last\-column> \- the number or name of the last column to include
.IP
[\-reverse] \- use the range in reverse order
.IP
Takes one or more metric files and constructs a new metric file by
concatenating columns from them.  The input metric files must have the
same number of vertices and same structure.
.IP
Example: wb_command \fB\-metric\-merge\fR out.func.gii \fB\-metric\fR first.func.gii
\fB\-column\fR 1 \fB\-metric\fR second.func.gii
.IP
This example would take the first column from first.func.gii, followed by
all columns from second.func.gii, and write these columns to
out.func.gii.
.PP
\fB\-metric\-palette\fR
SET THE PALETTE OF A METRIC FILE
.IP
wb_command \fB\-metric\-palette\fR
.IP
<metric> \- the metric to modify
<mode> \- the mapping mode
.IP
[\-column] \- select a single column
.IP
<column> \- the column number or name
.IP
[\-pos\-percent] \- percentage min/max for positive data coloring
.IP
<pos\-min\-%> \- the percentile for the least positive data
<pos\-max\-%> \- the percentile for the most positive data
.IP
[\-neg\-percent] \- percentage min/max for negative data coloring
.IP
<neg\-min\-%> \- the percentile for the least negative data
<neg\-max\-%> \- the percentile for the most negative data
.IP
[\-pos\-user] \- user min/max values for positive data coloring
.IP
<pos\-min\-user> \- the value for the least positive data
<pos\-max\-user> \- the value for the most positive data
.IP
[\-neg\-user] \- user min/max values for negative data coloring
.IP
<neg\-min\-user> \- the value for the least negative data
<neg\-max\-user> \- the value for the most negative data
.IP
[\-interpolate] \- interpolate colors
.IP
<interpolate> \- boolean, whether to interpolate
.IP
[\-disp\-pos] \- display positive data
.IP
<display> \- boolean, whether to display
.IP
[\-disp\-neg] \- display positive data
.IP
<display> \- boolean, whether to display
.IP
[\-disp\-zero] \- display data closer to zero than the min cutoff
.IP
<display> \- boolean, whether to display
.IP
[\-palette\-name] \- set the palette used
.IP
<name> \- the name of the palette
.IP
[\-thresholding] \- set the thresholding
.IP
<type> \- thresholding setting
<test> \- show values inside or outside thresholds
<min> \- lower threshold
<max> \- upper threshold
.TP
The original metric file is overwritten with the modified version.
By
.IP
default, all columns of the metric file are adjusted to the new settings,
use the \fB\-column\fR option to change only one column.  Mapping settings not
specified in options will be taken from the first column.  The <mode>
argument must be one of the following:
.IP
MODE_AUTO_SCALE
MODE_AUTO_SCALE_ABSOLUTE_PERCENTAGE
MODE_AUTO_SCALE_PERCENTAGE
MODE_USER_SCALE
.IP
The <name> argument to \fB\-palette\-name\fR must be one of the following:
.IP
PSYCH
PSYCH\-NO\-NONE
ROY\-BIG
ROY\-BIG\-BL
Orange\-Yellow
Gray_Interp_Positive
Gray_Interp
clear_brain
videen_style
fidl
raich4_clrmid
raich6_clrmid
HSB8_clrmid
RBGYR20
RBGYR20P
POS_NEG
red\-yellow
blue\-lightblue
FSL
power_surf
fsl_red
fsl_green
fsl_blue
fsl_yellow
JET256
.IP
The <type> argument to \fB\-thresholding\fR must be one of the following:
.IP
THRESHOLD_TYPE_OFF
THRESHOLD_TYPE_NORMAL
.IP
The <test> argument to \fB\-thresholding\fR must be one of the following:
.IP
THRESHOLD_TEST_SHOW_OUTSIDE
THRESHOLD_TEST_SHOW_INSIDE
.PP
\fB\-metric\-reduce\fR
PERFORM REDUCTION OPERATION ACROSS METRIC COLUMNS
.IP
wb_command \fB\-metric\-reduce\fR
.IP
<metric\-in> \- the metric to reduce
<operation> \- the reduction operator to use
<metric\-out> \- output \- the output metric
.IP
[\-exclude\-outliers] \- exclude non\-numeric values and outliers by standard
.IP
deviation
<sigma\-below> \- number of standard deviations below the mean to
.IP
include
.IP
<sigma\-above> \- number of standard deviations above the mean to
.IP
include
.IP
[\-only\-numeric] \- exclude non\-numeric values
.IP
For each surface vertex, takes the data across columns as a vector, and
performs the specified reduction on it, putting the result into the
single output column at that vertex.  The reduction operators are as
follows:
.IP
MAX: the maximum value
MIN: the minimum value
INDEXMAX: the 1\-based index of the maximum value
INDEXMIN: the 1\-based index of the minimum value
SUM: add all values
PRODUCT: multiply all values
MEAN: the mean of the data
STDEV: the standard deviation (N denominator)
SAMPSTDEV: the sample standard deviation (N\-1 denominator)
VARIANCE: the variance of the data
TSNR: mean divided by sample standard deviation (N\-1 denominator)
COV: sample standard deviation (N\-1 denominator) divided by mean
MEDIAN: the median of the data
MODE: the mode of the data
COUNT_NONZERO: the number of nonzero elements in the data
.PP
\fB\-metric\-regression\fR
REGRESS METRICS OUT OF A METRIC FILE
.IP
wb_command \fB\-metric\-regression\fR
.IP
<metric\-in> \- the metric to regress from
<metric\-out> \- output \- the output metric
.IP
[\-roi] \- only regress inside an roi
.IP
<roi\-metric> \- the area to use for regression, as a metric
.IP
[\-column] \- select a single column to regress from
.IP
<column> \- the column number or name
.IP
[\-remove] \- repeatable \- specify a metric to regress out
.IP
<metric> \- the metric file to use
.IP
[\-remove\-column] \- select a column to use, rather than all
.IP
<column> \- the column number or name
.IP
[\-keep] \- repeatable \- specify a metric to include in regression, but not
.IP
remove
<metric> \- the metric file to use
.IP
[\-keep\-column] \- select a column to use, rather than all
.IP
<column> \- the column number or name
.IP
For each regressor, its mean across the surface is subtracted from its
data.  Each input map is then regressed against these, and a constant
term.  The resulting regressed slopes of all regressors specified with
\fB\-remove\fR are multiplied with their respective regressor maps, and these
are subtracted from the input map.
.PP
\fB\-metric\-remove\-islands\fR
REMOVE ISLANDS FROM AN ROI METRIC
.IP
wb_command \fB\-metric\-remove\-islands\fR
.IP
<surface> \- the surface to use for neighbor information
<metric\-in> \- the input ROI metric
<metric\-out> \- output \- the output ROI metric
.IP
[\-corrected\-areas] \- vertex areas to use instead of computing them from
.IP
the surface
<area\-metric> \- the corrected vertex areas, as a metric
.IP
Finds all connected areas in the ROI, and zeros out all but the largest
one, in terms of surface area.
.PP
\fB\-metric\-resample\fR
RESAMPLE A METRIC FILE TO A DIFFERENT MESH
.IP
wb_command \fB\-metric\-resample\fR
.IP
<metric\-in> \- the metric file to resample
<current\-sphere> \- a sphere surface with the mesh that the metric is
.IP
currently on
.IP
<new\-sphere> \- a sphere surface that is in register with <current\-sphere>
.IP
and has the desired output mesh
.IP
<method> \- the method name
<metric\-out> \- output \- the output metric
.IP
[\-area\-surfs] \- specify surfaces to do vertex area correction based on
.IP
<current\-area> \- a relevant anatomical surface with <current\-sphere>
.IP
mesh
.IP
<new\-area> \- a relevant anatomical surface with <new\-sphere> mesh
.IP
[\-area\-metrics] \- specify vertex area metrics to do area correction based
.IP
on
<current\-area> \- a metric file with vertex areas for <current\-sphere>
.IP
mesh
.IP
<new\-area> \- a metric file with vertex areas for <new\-sphere> mesh
.IP
[\-current\-roi] \- use an input roi on the current mesh to exclude non\-data
.IP
vertices
<roi\-metric> \- the roi, as a metric file
.IP
[\-valid\-roi\-out] \- output the ROI of vertices that got data from valid
.IP
source vertices
<roi\-out> \- output \- the output roi as a metric
.IP
[\-largest] \- use only the value of the vertex with the largest weight
.IP
Resamples a metric file, given two spherical surfaces that are in
register.  If ADAP_BARY_AREA is used, exactly one of \fB\-area\-surfs\fR or
\fB\-area\-metrics\fR must be specified.
.IP
The ADAP_BARY_AREA method is recommended for ordinary metric data,
because it should use all data while downsampling, unlike BARYCENTRIC.
The recommended areas option for most data is individual midthicknesses
for individual data, and averaged vertex area metrics from individual
midthicknesses for group average data.
.IP
The \fB\-current\-roi\fR option only masks the input, the output may be slightly
dilated in comparison, consider using \fB\-metric\-mask\fR on the output when
using \fB\-current\-roi\fR.
.IP
The \fB\-largest\fR option results in nearest vertex behavior when used with
BARYCENTRIC.  When resampling a binary metric, consider thresholding at
0.5 after resampling rather than using \fB\-largest\fR.
.IP
The <method> argument must be one of the following:
.IP
ADAP_BARY_AREA
BARYCENTRIC
.PP
\fB\-metric\-rois\-from\-extrema\fR
CREATE METRIC ROI MAPS FROM EXTREMA MAPS
.IP
wb_command \fB\-metric\-rois\-from\-extrema\fR
.IP
<surface> \- the surface to use for geodesic distance
<metric> \- the input metric file
<limit> \- geodesic distance limit from vertex, in mm
<metric\-out> \- output \- the output metric file
.IP
[\-gaussian] \- generate a gaussian kernel instead of a flat ROI
.IP
<sigma> \- the sigma for the gaussian kernel, in mm
.IP
[\-roi] \- select a region of interest to use
.IP
<roi\-metric> \- the area to use, as a metric
.IP
[\-overlap\-logic] \- how to handle overlapping ROIs, default ALLOW
.IP
<method> \- the method of resolving overlaps
.IP
[\-column] \- select a single input column to use
.IP
<column> \- the column number or name
.IP
For each nonzero value in each map, make a map with an ROI around that
location.  If the \fB\-gaussian\fR option is specified, then normalized gaussian
kernels are output instead of ROIs.  The <method> argument to
\fB\-overlap\-logic\fR must be one of ALLOW, CLOSEST, or EXCLUDE.  ALLOW is the
default, and means that ROIs are treated independently and may overlap.
CLOSEST means that ROIs may not overlap, and that no ROI contains
vertices that are closer to a different seed vertex.  EXCLUDE means that
ROIs may not overlap, and that any vertex within range of more than one
ROI does not belong to any ROI.
.PP
\fB\-metric\-rois\-to\-border\fR
DRAW BORDERS AROUND METRIC ROIS
.IP
wb_command \fB\-metric\-rois\-to\-border\fR
.IP
<surface> \- the surface to use for neighbor information
<metric> \- the input metric containing ROIs
<class\-name> \- the name to use for the class of the output borders
<border\-out> \- output \- the output border file
.IP
[\-placement] \- set how far along the edge border points are drawn
.IP
<fraction> \- fraction along edge from inside vertex (default 0.33)
.IP
[\-column] \- select a single column
.IP
<column> \- the column number or name
.IP
For each ROI column, finds all edges on the mesh that cross the boundary
of the ROI, and draws borders through them.  By default, this is done on
all columns in the input file, using the map name as the name for the
border.
.PP
\fB\-metric\-smoothing\fR
SMOOTH A METRIC FILE
.IP
wb_command \fB\-metric\-smoothing\fR
.IP
<surface> \- the surface to smooth on
<metric\-in> \- the metric to smooth
<smoothing\-kernel> \- the sigma for the gaussian kernel function, in mm
<metric\-out> \- output \- the output metric
.IP
[\-roi] \- select a region of interest to smooth
.IP
<roi\-metric> \- the roi to smooth within, as a metric
.IP
[\-match\-columns] \- for each input column, use the corresponding column
.IP
from the roi
.IP
[\-fix\-zeros] \- treat zero values as not being data
.IP
[\-column] \- select a single column to smooth
.IP
<column> \- the column number or name
.IP
[\-corrected\-areas] \- vertex areas to use instead of computing them from
.IP
the surface
<area\-metric> \- the corrected vertex areas, as a metric
.IP
[\-method] \- select smoothing method, default GEO_GAUSS_AREA
.IP
<method> \- the name of the smoothing method
.TP
Smooth a metric file on a surface.
By default, smooths all input columns
.IP
on the entire surface, specify \fB\-column\fR to use only one input column, and
\fB\-roi\fR to smooth only where the roi metric is greater than 0, outputting
zeros elsewhere.
.IP
When using \fB\-roi\fR, input data outside the ROI is not used to compute the
smoothed values.  By default, the first column of the roi metric is used
for all input columns.  When \fB\-match\-columns\fR is specified to the \fB\-roi\fR
option, the input and roi metrics must have the same number of columns,
and for each input column's index, the same column index is used in the
roi metric.  If the \fB\-match\-columns\fR option to \fB\-roi\fR is used while the
\fB\-column\fR option is also used, the number of columns must match between the
roi and input metric, and it will use the roi column with the index of
the selected input column.
.IP
The \fB\-fix\-zeros\fR option causes the smoothing to not use an input value if
it is zero, but still write a smoothed value to the vertex.  This is
useful for zeros that indicate lack of information, preventing them from
pulling down the intensity of nearby vertices, while giving the zero an
extrapolated value.
.IP
The \fB\-corrected\-areas\fR option is intended for when it is unavoidable to
smooth on a group average surface, it is only an approximate correction
for the reduction of structure in a group average surface.  It is better
to smooth the data on individuals before averaging, when feasible.
.IP
Valid values for <method> are:
.IP
GEO_GAUSS_AREA \- uses a geodesic gaussian kernel, and normalizes based on
vertex area in order to work more reliably on irregular surfaces
.IP
GEO_GAUSS_EQUAL \- uses a geodesic gaussian kernel, and normalizes
assuming each vertex has equal importance
.IP
GEO_GAUSS \- matches geodesic gaussian smoothing from caret5, but does not
check kernels for having unequal importance
.IP
The GEO_GAUSS_AREA method is the default because it is usually the
correct choice.  GEO_GAUSS_EQUAL may be the correct choice when the sum
of vertex values is more meaningful then the surface integral (sum of
values .* areas), for instance when smoothing vertex areas (the sum is
the total surface area, while the surface integral is the sum of squares
of the vertex areas).  The GEO_GAUSS method is not recommended, it exists
mainly to replicate methods of studies done with caret5's geodesic
smoothing.
.PP
\fB\-metric\-stats\fR
SPATIAL STATISTICS ON A METRIC FILE
.IP
wb_command \fB\-metric\-stats\fR
.IP
<metric\-in> \- the input metric
.IP
[\-reduce] \- use a reduction operation
.IP
<operation> \- the reduction operation
.IP
[\-percentile] \- give the value at a percentile
.IP
<percent> \- the percentile to find
.IP
[\-column] \- only display output for one column
.IP
<column> \- the column number or name
.IP
[\-roi] \- only consider data inside an roi
.IP
<roi\-metric> \- the roi, as a metric file
.IP
[\-match\-maps] \- each column of input uses the corresponding column
.IP
from the roi file
.IP
[\-show\-map\-name] \- print map index and name before each output
.IP
For each column of the input, a single number is printed, resulting from
the specified reduction or percentile operation.  Use \fB\-column\fR to only
give output for a single column.  Use \fB\-roi\fR to consider only the data
within a region.  Exactly one of \fB\-reduce\fR or \fB\-percentile\fR must be
specified.
.IP
The argument to the \fB\-reduce\fR option must be one of the following:
.IP
MAX: the maximum value
MIN: the minimum value
INDEXMAX: the 1\-based index of the maximum value
INDEXMIN: the 1\-based index of the minimum value
SUM: add all values
PRODUCT: multiply all values
MEAN: the mean of the data
STDEV: the standard deviation (N denominator)
SAMPSTDEV: the sample standard deviation (N\-1 denominator)
VARIANCE: the variance of the data
TSNR: mean divided by sample standard deviation (N\-1 denominator)
COV: sample standard deviation (N\-1 denominator) divided by mean
MEDIAN: the median of the data
MODE: the mode of the data
COUNT_NONZERO: the number of nonzero elements in the data
.PP
\fB\-metric\-tfce\fR
DO TFCE ON A METRIC FILE
.IP
wb_command \fB\-metric\-tfce\fR
.IP
<surface> \- the surface to compute on
<metric\-in> \- the metric to run TFCE on
<metric\-out> \- output \- the output metric
.IP
[\-presmooth] \- smooth the metric before running TFCE
.IP
<kernel> \- the sigma for the gaussian smoothing kernel, in mm
.IP
[\-roi] \- select a region of interest to run TFCE on
.IP
<roi\-metric> \- the area to run TFCE on, as a metric
.IP
[\-parameters] \- set parameters for TFCE integral
.IP
<E> \- exponent for cluster area (default 1.0)
<H> \- exponent for threshold value (default 2.0)
.IP
[\-column] \- select a single column
.IP
<column> \- the column number or name
.IP
[\-corrected\-areas] \- vertex areas to use instead of computing them from
.IP
the surface
<area\-metric> \- the corrected vertex areas, as a metric
.IP
Threshold\-free cluster enhancement is a method to increase the relative
value of regions that would form clusters in a standard thresholding
test.  This is accomplished by evaluating the integral of:
.IP
e(h, p)^E * h^H * dh
.IP
at each vertex p, where h ranges from 0 to the maximum value in the data,
and e(h, p) is the extent of the cluster containing vertex p at threshold
h.  Negative values are similarly enhanced by negating the data, running
the same process, and negating the result.
.IP
When using \fB\-presmooth\fR with \fB\-corrected\-areas\fR, note that it is an
approximate correction within the smoothing algorithm (the TFCE
correction is exact).  Doing smoothing on individual surfaces before
averaging/TFCE is preferred, when possible, in order to better tie the
smoothing kernel size to the original feature size.
.IP
The TFCE method is explained in: Smith SM, Nichols TE., "Threshold\-free
cluster enhancement: addressing problems of smoothing, threshold
dependence and localisation in cluster inference." Neuroimage. 2009 Jan
1;44(1):83\-98. PMID: 18501637
.PP
\fB\-metric\-to\-volume\-mapping\fR
MAP METRIC FILE TO VOLUME
.IP
wb_command \fB\-metric\-to\-volume\-mapping\fR
.IP
<metric> \- the input metric file
<surface> \- the surface to use coordinates from
<volume\-space> \- a volume file in the desired output volume space
<volume\-out> \- output \- the output volume file
.IP
[\-nearest\-vertex] \- use the value from the vertex closest to the voxel
.IP
center
<distance> \- how far from the surface to map values to voxels, in mm
.IP
[\-ribbon\-constrained] \- use ribbon constrained mapping algorithm
.IP
<inner\-surf> \- the inner surface of the ribbon
<outer\-surf> \- the outer surface of the ribbon
.IP
[\-voxel\-subdiv] \- voxel divisions while estimating voxel weights
.IP
<subdiv\-num> \- number of subdivisions, default 3
.TP
Maps values from a metric file into a volume file.
You must specify
.TP
exactly one mapping method option.
The \fB\-nearest\-vertex\fR method uses the
.IP
value from the vertex closest to the voxel center (useful for integer
values).  The \fB\-ribbon\-constrained\fR method uses the same method as in
\fB\-volume\-to\-surface\-mapping\fR, then uses the weights in reverse.
.PP
\fB\-metric\-vector\-operation\fR
DO A VECTOR OPERATION ON METRIC FILES
.IP
wb_command \fB\-metric\-vector\-operation\fR
.IP
<vectors\-a> \- first vector input file
<vectors\-b> \- second vector input file
<operation> \- what vector operation to do
<metric\-out> \- output \- the output file
.IP
[\-normalize\-a] \- normalize vectors of first input
.IP
[\-normalize\-b] \- normalize vectors of second input
.IP
[\-normalize\-output] \- normalize output vectors (not valid for dot
.IP
product)
.IP
[\-magnitude] \- output the magnitude of the result (not valid for dot
.IP
product)
.IP
Does a vector operation on two metric files (that must have a multiple of
3 columns).  Either of the inputs may have multiple vectors (more than 3
columns), but not both (at least one must have exactly 3 columns).  The
\fB\-magnitude\fR and \fB\-normalize\-output\fR options may not be specified together,
or with an operation that returns a scalar (dot product).  The
<operation> parameter must be one of the following:
.IP
DOT
CROSS
ADD
SUBTRACT
.PP
\fB\-metric\-vector\-toward\-roi\fR
FIND IF VECTORS POINT TOWARD AN ROI
.IP
wb_command \fB\-metric\-vector\-toward\-roi\fR
.IP
<surface> \- the surface to compute on
<target\-roi> \- the roi to find the shortest path to
<metric\-out> \- output \- the output metric
.IP
[\-roi] \- don't compute for vertices outside an roi
.IP
<roi\-metric> \- the region to compute inside, as a metric
.IP
At each vertex, compute the vector along the start of the shortest path
to the ROI.
.PP
\fB\-metric\-weighted\-stats\fR
WEIGHTED SPATIAL STATISTICS ON A METRIC FILE
.IP
wb_command \fB\-metric\-weighted\-stats\fR
.IP
<metric\-in> \- the input metric
.IP
[\-area\-surface] \- use vertex areas as weights
.IP
<area\-surface> \- the surface to use for vertex areas
.IP
[\-weight\-metric] \- use weights from a metric file
.IP
<weight\-metric> \- metric file containing the weights
.IP
[\-column] \- only display output for one column
.IP
<column> \- the column number or name
.IP
[\-roi] \- only consider data inside an roi
.IP
<roi\-metric> \- the roi, as a metric file
.IP
[\-match\-maps] \- each column of input uses the corresponding column
.IP
from the roi file
.IP
[\-mean] \- compute weighted mean
.IP
[\-stdev] \- compute weighted standard deviation
.IP
[\-sample] \- estimate population stdev from the sample
.IP
[\-percentile] \- compute weighted percentile
.IP
<percent> \- the percentile to find
.IP
[\-sum] \- compute weighted sum
.IP
[\-show\-map\-name] \- print map index and name before each output
.IP
For each column of the input, a single number is printed, resulting from
the specified operation.  You must specify exactly one of \fB\-area\-surface\fR
or \fB\-weight\-metric\fR.  Use \fB\-column\fR to only give output for a single column.
Use \fB\-roi\fR to consider only the data within a region.  Exactly one of
\fB\-mean\fR, \fB\-stdev\fR, \fB\-percentile\fR or \fB\-sum\fR must be specified.
.IP
Using \fB\-sum\fR with \fB\-area\-surface\fR (or \fB\-weight\-metric\fR with a metric containing
similar data) is equivalent to integrating with respect to surface area.
For example, if you want to find the surface area within an roi, do this:
.IP
\f(CW$ wb_command -metric-weighted-stats roi.func.gii -sum -area-surface\fR
.IP
midthickness.surf.gii
.PP
\fB\-nifti\-information\fR
DISPLAY INFORMATION ABOUT A NIFTI/CIFTI FILE
.IP
wb_command \fB\-nifti\-information\fR
.IP
<nifti\-file> \- the nifti/cifti file to examine
.IP
[\-print\-header] \- display the header contents
.IP
[\-print\-matrix] \- output the values in the matrix (cifti only)
.IP
[\-print\-xml] \- print the cifti XML (cifti only)
.IP
[\-version] \- convert the XML to a specific CIFTI version (default is
.IP
the file's cifti version)
<version> \- the CIFTI version to use
.IP
You must specify at least one \fB\-print\-\fR* option.
.PP
\fB\-probtrackx\-dot\-convert\fR
CONVERT A .DOT FILE FROM PROBTRACKX TO CIFTI
.IP
wb_command \fB\-probtrackx\-dot\-convert\fR
.IP
<dot\-file> \- input .dot file
<cifti\-out> \- output \- output cifti file
.IP
[\-row\-voxels] \- the output mapping along a row will be voxels
.IP
<voxel\-list\-file> \- a text file containing IJK indices for the voxels
.IP
used
.IP
<label\-vol> \- a label volume with the dimensions and sform used, with
.IP
structure labels
.IP
[\-row\-surface] \- the output mapping along a row will be surface vertices
.IP
<roi\-metric> \- a metric file with positive values on all vertices used
.IP
[\-row\-cifti] \- take the mapping along a row from a cifti file
.IP
<cifti> \- the cifti file to take the mapping from
<direction> \- which dimension to take the mapping along, ROW or COLUMN
.IP
[\-col\-voxels] \- the output mapping along a column will be voxels
.IP
<voxel\-list\-file> \- a text file containing IJK indices for the voxels
.IP
used
.IP
<label\-vol> \- a label volume with the dimensions and sform used, with
.IP
structure labels
.IP
[\-col\-surface] \- the output mapping along a column will be surface
.IP
vertices
<roi\-metric> \- a metric file with positive values on all vertices used
.IP
[\-col\-cifti] \- take the mapping along a column from a cifti file
.IP
<cifti> \- the cifti file to take the mapping from
<direction> \- which dimension to take the mapping along, ROW or COLUMN
.IP
[\-transpose] \- transpose the input matrix
.IP
[\-make\-symmetric] \- transform half\-square input into full matrix output
.IP
NOTE: exactly one \fB\-row\fR option and one \fB\-col\fR option must be used.
.IP
If the input file does not have its indexes sorted in the correct
ordering, this command may take longer than expected.  Specifying
\fB\-transpose\fR will transpose the input matrix before trying to put its
values into the cifti file, which is currently needed for at least
matrix2 in order to display it as intended.  How the cifti file is
displayed is based on which \fB\-row\fR option is specified: if \fB\-row\-voxels\fR is
specified, then it will display data on volume slices.  The label names
in the label volume(s) must have the following names, other names are
ignored:
.IP
CORTEX_LEFT
CORTEX_RIGHT
CEREBELLUM
ACCUMBENS_LEFT
ACCUMBENS_RIGHT
ALL_GREY_MATTER
ALL_WHITE_MATTER
AMYGDALA_LEFT
AMYGDALA_RIGHT
BRAIN_STEM
CAUDATE_LEFT
CAUDATE_RIGHT
CEREBELLAR_WHITE_MATTER_LEFT
CEREBELLAR_WHITE_MATTER_RIGHT
CEREBELLUM_LEFT
CEREBELLUM_RIGHT
CEREBRAL_WHITE_MATTER_LEFT
CEREBRAL_WHITE_MATTER_RIGHT
CORTEX
DIENCEPHALON_VENTRAL_LEFT
DIENCEPHALON_VENTRAL_RIGHT
HIPPOCAMPUS_LEFT
HIPPOCAMPUS_RIGHT
INVALID
OTHER
OTHER_GREY_MATTER
OTHER_WHITE_MATTER
PALLIDUM_LEFT
PALLIDUM_RIGHT
PUTAMEN_LEFT
PUTAMEN_RIGHT
THALAMUS_LEFT
THALAMUS_RIGHT
.PP
\fB\-scene\-file\-merge\fR
REARRANGE SCENES INTO A NEW FILE
.IP
wb_command \fB\-scene\-file\-merge\fR
.IP
<scene\-file\-out> \- output \- the output scene file
.IP
[\-scene\-file] \- repeatable \- specify a scene file to use scenes from
.IP
<scene\-file> \- the input scene file
.IP
[\-scene] \- repeatable \- specify a scene to use
.IP
<scene> \- the scene number or name
.IP
[\-up\-to] \- use an inclusive range of scenes
.IP
<last\-column> \- the number or name of the last scene to include
.IP
[\-reverse] \- use the range in reverse order
.IP
Takes one or more scene files and constructs a new scene file by
concatenating specified scenes from them.
.IP
Example: wb_command \fB\-scene\-file\-merge\fR out.scene \fB\-scene\-file\fR first.scene
\fB\-scene\fR 1 \fB\-scene\-file\fR second.scene
.IP
This example would take the first scene from first.scene, followed by all
scenes from second.scene, and write these scenes to out.scene.
.PP
\fB\-scene\-file\-relocate\fR
RECREATE SCENE FILE IN NEW LOCATION
.IP
wb_command \fB\-scene\-file\-relocate\fR
.IP
<input\-scene> \- the scene file to use
<output\-scene> \- output \- the new scene file to create
.IP
Scene files contain internal relative paths, such that moving or copying
a scene file will cause it to lose track of the files it refers to.  This
command makes a modified copy of the scene file, changing the relative
paths to refer to the new relative locations of the files.
.PP
\fB\-set\-map\-names\fR
SET THE NAME OF ONE OR MORE MAPS IN A FILE
.IP
wb_command \fB\-set\-map\-names\fR
.IP
<data\-file> \- the file to set the map names of
.IP
[\-name\-file] \- use a text file to replace all map names
.IP
<file> \- text file containing map names, one per line
.IP
[\-map] \- repeatable \- specify a map to set the name of
.IP
<index> \- the map index to change the name of
<new\-name> \- the name to set for the map
.IP
Sets the name of one or more maps for metric, shape, label, volume, cifti
scalar or cifti label files.  If the \fB\-name\-file\fR option is not specified,
the \fB\-map\fR option must be specified at least once.  The \fB\-map\fR option cannot
be used when \fB\-name\-file\fR is specified.
.PP
\fB\-set\-structure\fR
SET STRUCTURE OF A DATA FILE
.IP
wb_command \fB\-set\-structure\fR
.IP
<data\-file> \- the file to set the structure of
<structure> \- the structure to set the file to
.IP
[\-surface\-type] \- set the type of a surface (only used if file is a
.IP
surface file)
<type> \- name of surface type
.IP
[\-surface\-secondary\-type] \- set the secondary type of a surface (only
.IP
used if file is a surface file)
<secondary type> \- name of surface secondary type
.TP
The existing file is modified and rewritten to the same filename.
Valid
.IP
values for the structure name are:
.IP
CORTEX_LEFT
CORTEX_RIGHT
CEREBELLUM
ACCUMBENS_LEFT
ACCUMBENS_RIGHT
ALL_GREY_MATTER
ALL_WHITE_MATTER
AMYGDALA_LEFT
AMYGDALA_RIGHT
BRAIN_STEM
CAUDATE_LEFT
CAUDATE_RIGHT
CEREBELLAR_WHITE_MATTER_LEFT
CEREBELLAR_WHITE_MATTER_RIGHT
CEREBELLUM_LEFT
CEREBELLUM_RIGHT
CEREBRAL_WHITE_MATTER_LEFT
CEREBRAL_WHITE_MATTER_RIGHT
CORTEX
DIENCEPHALON_VENTRAL_LEFT
DIENCEPHALON_VENTRAL_RIGHT
HIPPOCAMPUS_LEFT
HIPPOCAMPUS_RIGHT
INVALID
OTHER
OTHER_GREY_MATTER
OTHER_WHITE_MATTER
PALLIDUM_LEFT
PALLIDUM_RIGHT
PUTAMEN_LEFT
PUTAMEN_RIGHT
THALAMUS_LEFT
THALAMUS_RIGHT
.IP
Valid names for the surface type are:
.IP
UNKNOWN
RECONSTRUCTION
ANATOMICAL
INFLATED
VERY_INFLATED
SPHERICAL
SEMI_SPHERICAL
ELLIPSOID
FLAT
HULL
.IP
Valid names for the surface secondary type are:
.IP
INVALID
GRAY_WHITE
MIDTHICKNESS
PIAL
.PP
\fB\-show\-scene\fR
OFFSCREEN RENDERING OF SCENE TO AN IMAGE FILE
.IP
wb_command \fB\-show\-scene\fR
.IP
<scene\-file> \- scene file
<scene\-name\-or\-number> \- name or number (starting at one) of the scene in
.IP
the scene file
.IP
<image\-file\-name> \- output image file name
<image\-width> \- width of output image(s)
<image\-height> \- height of output image(s)
.IP
[\-use\-window\-size] \- Override image size with window size
.IP
[\-no\-scene\-colors] \- Do not use background and foreground colors in scene
.IP
Render content of browser windows displayed in a scene into image
file(s).  The image file name should be similar to "capture.png".  If
there is only one image to render, the image name will not change.  If
there is more than one image to render, an index will be inserted into
the image name: "capture_01.png", "capture_02.png" etc.
.IP
The image format is determined by the image file extension.
The available image formats may vary by operating system.
Image formats available on this system are:
.IP
bmp
ico
jpeg
jpg
png
ppm
tif
tiff
xbm
xpm
.IP
The result of using the "\-use\-window\-size" option
is dependent upon the version used to create the scene.
.IP
* Versions 1.2 and newer contain the width and
.TP
height of the graphics region.
The output image
.IP
will be the width and height from the scene and
the image width and height specified on the command
line is ignored.
.IP
* If the scene does not contain the width and height
.IP
of the graphics region, the width and height specified
on the command line is used for the size of the
output image.
.PP
\fB\-signed\-distance\-to\-surface\fR
COMPUTE SIGNED DISTANCE FROM ONE SURFACE TO ANOTHER
.IP
wb_command \fB\-signed\-distance\-to\-surface\fR
.IP
<surface\-comp> \- the comparison surface to measure the signed distance on
<surface\-ref> \- the reference surface that defines the signed distance
.IP
function
.IP
<metric> \- output \- the output metric
.IP
[\-winding] \- winding method for point inside surface test
.IP
<method> \- name of the method (default EVEN_ODD)
.IP
Compute the signed distance function of the reference surface at every
vertex on the comparison surface.  NOTE: this relation is NOT symmetric,
the line from a vertex to the closest point on the 'ref' surface (the one
that defines the signed distance function) will only align with the
normal of the 'ref' surface.  Valid specifiers for winding methods are as
follows:
.IP
EVEN_ODD (default)
NEGATIVE
NONZERO
NORMALS
.IP
The NORMALS method uses the normals of triangles and edges, or the
closest triangle hit by a ray from the point.  This method may be
slightly faster, but is only reliable for a closed surface that does not
cross through itself.  All other methods count entry (positive) and exit
(negative) crossings of a vertical ray from the point, then counts as
inside if the total is odd, negative, or nonzero, respectively.
.PP
\fB\-spec\-file\-merge\fR
MERGE TWO SPEC FILES INTO ONE
.IP
wb_command \fB\-spec\-file\-merge\fR
.IP
<spec\-1> \- first spec file to merge
<spec\-2> \- second spec file to merge
<out\-spec> \- output \- output spec file
.IP
The output spec file contains every file that is in either of the input
spec files.
.PP
\fB\-spec\-file\-relocate\fR
RECREATE SPEC FILE IN NEW LOCATION
.IP
wb_command \fB\-spec\-file\-relocate\fR
.IP
<input\-spec> \- the spec file to use
<output\-spec> \- output \- the new spec file to create
.IP
Spec files contain internal relative paths, such that moving or copying a
spec file will cause it to lose track of the files it refers to.  This
command makes a modified copy of the spec file, changing the relative
paths to refer to the new relative locations of the files.
.PP
\fB\-surface\-affine\-regression\fR
REGRESS THE AFFINE TRANSFORM BETWEEN SURFACES ON THE SAME MESH
.IP
wb_command \fB\-surface\-affine\-regression\fR
.IP
<source> \- the surface to warp
<target> \- the surface to match the coordinates of
<affine\-out> \- output \- the output affine file
.IP
Use linear regression to compute an affine that minimizes the sum of
squares of the coordinate differences between the target surface and the
warped source surface.  Note that this has a bias to shrink the surface
that is being warped.  The output is written as a NIFTI 'world' matrix,
see \fB\-convert\-affine\fR to convert it for use in other software.
.PP
\fB\-surface\-apply\-affine\fR
APPLY AFFINE TRANSFORM TO SURFACE FILE
.IP
wb_command \fB\-surface\-apply\-affine\fR
.IP
<in\-surf> \- the surface to transform
<affine> \- the affine file
<out\-surf> \- output \- the output transformed surface
.IP
[\-flirt] \- MUST be used if affine is a flirt affine
.IP
<source\-volume> \- the source volume used when generating the affine
<target\-volume> \- the target volume used when generating the affine
.IP
For flirt matrices, you must use the \fB\-flirt\fR option, because flirt
matrices are not a complete description of the coordinate transform they
represent.  If the \fB\-flirt\fR option is not present, the affine must be a
nifti 'world' affine, which can be obtained with the \fB\-convert\-affine\fR
command, or aff_conv from the 4dfp suite.
.PP
\fB\-surface\-apply\-warpfield\fR
APPLY WARPFIELD TO SURFACE FILE
.IP
wb_command \fB\-surface\-apply\-warpfield\fR
.IP
<in\-surf> \- the surface to transform
<warpfield> \- the INVERSE warpfield
<out\-surf> \- output \- the output transformed surface
.IP
[\-fnirt] \- MUST be used if using a fnirt warpfield
.IP
<forward\-warp> \- the forward warpfield
.IP
NOTE: warping a surface requires the INVERSE of the warpfield used to
warp the volume it lines up with.  The header of the forward warp is
needed by the \fB\-fnirt\fR option in order to correctly interpret the
displacements in the fnirt warpfield.
.IP
If the \fB\-fnirt\fR option is not present, the warpfield must be a nifti
\&'world' warpfield, which can be obtained with the \fB\-convert\-warpfield\fR
command.
.PP
\fB\-surface\-average\fR
AVERAGE SURFACE FILES TOGETHER
.IP
wb_command \fB\-surface\-average\fR
.IP
<surface\-out> \- output \- the output averaged surface
.IP
[\-stddev] \- compute 3D sample standard deviation
.IP
<stddev\-metric\-out> \- output \- the output metric for 3D sample
.IP
standard deviation
.IP
[\-uncertainty] \- compute caret5 'uncertainty'
.IP
<uncert\-metric\-out> \- output \- the output metric for uncertainty
.IP
[\-surf] \- repeatable \- specify a surface to include in the average
.IP
<surface> \- a surface file to average
.IP
[\-weight] \- specify a weighted average
.IP
<weight> \- the weight to use (default 1)
.IP
The 3D sample standard deviation is computed as
\&'sqrt(sum(squaredlength(xyz \- mean(xyz)))/(n \- 1))'.
.IP
Uncertainty is a legacy measure used in caret5, and is computed as
\&'sum(length(xyz \- mean(xyz)))/n'.
.IP
When weights are used, the 3D sample standard deviation treats them as
reliability weights.
.PP
\fB\-surface\-closest\-vertex\fR
FIND CLOSEST SURFACE VERTEX TO COORDINATES
.IP
wb_command \fB\-surface\-closest\-vertex\fR
.IP
<surface> \- the surface to use
<coord\-list\-file> \- text file with coordinates
<vertex\-list\-out> \- output \- the output text file with vertex numbers
.IP
For each coordinate XYZ triple, find the closest vertex in the surface,
and output its vertex number into a text file.  The input file should
only use whitespace to separate coordinates (spaces, newlines, tabs), for
instance:
.IP
20 30 25
30 \fB\-20\fR 10
.PP
\fB\-surface\-coordinates\-to\-metric\fR
MAKE METRIC FILE OF SURFACE COORDINATES
.IP
wb_command \fB\-surface\-coordinates\-to\-metric\fR
.IP
<surface> \- the surface to use the coordinates of
<metric\-out> \- output \- the output metric
.IP
Puts the coordinates of the surface into a 3\-map metric file, as x, y, z.
.PP
\fB\-surface\-cortex\-layer\fR
CREATE SURFACE APPROXIMATING A CORTICAL LAYER
.IP
wb_command \fB\-surface\-cortex\-layer\fR
.IP
<white\-surface> \- the white matter surface
<pial\-surface> \- the pial surface
<location> \- what volume fraction to place the layer at
<out\-surface> \- output \- the output surface
.IP
[\-placement\-out] \- output the placement as a distance fraction from pial
.IP
to white
<placement\-metric> \- output \- output metric
.IP
[\-untwist] \- temporary option for comparing methods, specify to use old
.IP
method
.TP
The input surfaces must have vertex correspondence.
The output surface
.IP
is generated by placing vertices between the two surfaces such that the
enclosed volume within any small patch of the new and white surfaces is
the given fraction of the volume of the same patch between the pial and
white surfaces (i.e., specifying 0 would give the white surface, 1 would
give the pial surface).
.PP
\fB\-surface\-create\-sphere\fR
GENERATE A SPHERE WITH CONSISTENT VERTEX AREAS
.IP
wb_command \fB\-surface\-create\-sphere\fR
.IP
<num\-vertices> \- desired number of vertices
<sphere\-out> \- output \- the output sphere
.IP
Generates a sphere by regularly dividing the triangles of an icosahedron,
to come as close to the desired number of vertices as possible, and
modifying it to have very similar vertex areas for all vertices.  To
generate a pair of vertex\-matched left and right spheres, use this
command, then \fB\-surface\-flip\-lr\fR to generate the other sphere, then
\fB\-set\-structure\fR on each.  For example:
.IP
\f(CW$ wb_command -surface-create-sphere 6000 Sphere.6k.R.surf.gii\fR
.br
\f(CW$ wb_command -surface-flip-lr Sphere.6k.R.surf.gii Sphere.6k.L.surf.gii\fR
.br
\f(CW$ wb_command -set-structure Sphere.6k.R.surf.gii CORTEX_RIGHT\fR
.br
\f(CW$ wb_command -set-structure Sphere.6k.L.surf.gii CORTEX_LEFT\fR
.PP
\fB\-surface\-curvature\fR
CALCULATE CURVATURE OF SURFACE
.IP
wb_command \fB\-surface\-curvature\fR
.IP
<surface> \- the surface to compute the curvature of
.IP
[\-mean] \- output mean curvature
.IP
<mean\-out> \- output \- mean curvature metric
.IP
[\-gauss] \- output gaussian curvature
.IP
<gauss\-out> \- output \- gaussian curvature metric
.IP
Compute the curvature of the surface, using the method from:
Interactive Texture Mapping by J. Maillot, Yahia, and Verroust, 1993.
ACM\-0\-98791\-601\-8/93/008
.PP
\fB\-surface\-cut\-resample\fR
RESAMPLE A CUT SURFACE
.IP
wb_command \fB\-surface\-cut\-resample\fR
.IP
<surface\-in> \- the surface file to resample
<current\-sphere> \- a sphere surface with the mesh that the input surface
.IP
is currently on
.IP
<new\-sphere> \- a sphere surface that is in register with <current\-sphere>
.IP
and has the desired output mesh
.IP
<surface\-out> \- output \- the output surface file
.IP
Resamples a surface file, given two spherical surfaces that are in
register.  Barycentric resampling is used, because it is usually better
for resampling surfaces, and because it is needed to figure out the new
topology anyway.
.PP
\fB\-surface\-distortion\fR
MEASURE DISTORTION BETWEEN SURFACES
.IP
wb_command \fB\-surface\-distortion\fR
.IP
<surface\-reference> \- the reference surface
<surface\-distorted> \- the distorted surface
<metric\-out> \- output \- the output distortion metric
.IP
[\-smooth] \- smooth the area data
.IP
<sigma> \- the smoothing kernel sigma in mm
.IP
[\-caret5\-method] \- use the surface distortion method from caret5
.IP
[\-edge\-method] \- calculate distortion of edge lengths rather than areas
.IP
This command, when not using \fB\-caret5\-method\fR or \fB\-edge\-method\fR, is
equivalent to using \fB\-surface\-vertex\-areas\fR on each surface, smoothing both
output metrics with the GEO_GAUSS_EQUAL method on the surface they came
from if \fB\-smooth\fR is specified, and then using the formula
\&'ln(distorted/reference)/ln(2)' on the smoothed results.
.IP
When using \fB\-caret5\-method\fR, it uses the surface distortion method from
caret5, which takes the base 2 log of the ratio of tile areas, then
averages those results at each vertex, and then smooths the result on the
reference surface.
.IP
When using \fB\-edge\-method\fR, the \fB\-smooth\fR option is ignored, and the output at
each vertex is the average of 'abs(ln(refEdge/distortEdge)/ln(2))' over
all edges connected to the vertex.
.PP
\fB\-surface\-flip\-lr\fR
MIRROR A SURFACE THROUGH THE YZ PLANE
.IP
wb_command \fB\-surface\-flip\-lr\fR
.IP
<surface> \- the surface to flip
<surface\-out> \- output \- the output flipped surface
.IP
This command negates the x coordinate of each vertex, and flips the
surface normals, so that you have a surface of opposite handedness with
the same features and vertex correspondence, with normals consistent with
the original surface.  That is, if the input surface has normals facing
outward, the output surface will also have normals facing outward.
.PP
\fB\-surface\-flip\-normals\fR
FLIP ALL TILES ON A SURFACE
.IP
wb_command \fB\-surface\-flip\-normals\fR
.IP
<surface> \- the surface to flip the normals of
<surface\-out> \- output \- the output surface
.IP
Flips all triangles on a surface, resulting in surface normals being
flipped the other direction (inward vs outward).  If you transform a
surface with an affine that has negative determinant, or a warpfield that
similarly flips the surface, you may end up with a surface that has
normals pointing inwards, which may have display problems.  Using this
command will solve that problem.
.PP
\fB\-surface\-generate\-inflated\fR
SURFACE GENERATE INFLATED
.IP
wb_command \fB\-surface\-generate\-inflated\fR
.IP
<anatomical\-surface\-in> \- the anatomical surface
<inflated\-surface\-out> \- output \- the output inflated surface
<very\-inflated\-surface\-out> \- output \- the output very inflated surface
.IP
[\-iterations\-scale] \- optional iterations scaling
.IP
<iterations\-scale\-value> \- iterations\-scale value
.IP
Generate inflated and very inflated surfaces. The output surfaces are
\&'matched' (have same XYZ range) to the anatomical surface. In most cases,
an iterations\-scale of 1.0 (default) is sufficient.  However, if the
surface contains a large number of vertices (150,000), try an
iterations\-scale of 2.5.
.PP
\fB\-surface\-geodesic\-distance\fR
COMPUTE GEODESIC DISTANCE FROM ONE VERTEX TO THE ENTIRE SURFACE
.IP
wb_command \fB\-surface\-geodesic\-distance\fR
.IP
<surface> \- the surface to compute on
<vertex> \- the vertex to compute geodesic distance from
<metric\-out> \- output \- the output metric
.IP
[\-naive] \- use only neighbors, don't crawl triangles (not recommended)
.IP
[\-limit] \- stop at a certain distance
.IP
<limit\-mm> \- distance in mm to stop at
.IP
Unless \fB\-limit\fR is specified, computes the geodesic distance from the
specified vertex to all others.  The result is output as a single column
metric file, with a value of \fB\-1\fR for vertices that the distance was not
computed for.  If \fB\-naive\fR is not specified, it uses not just immediate
neighbors, but also neighbors derived from crawling across pairs of
triangles that share an edge.
.PP
\fB\-surface\-geodesic\-rois\fR
DRAW GEODESIC LIMITED ROIS AT VERTICES
.IP
wb_command \fB\-surface\-geodesic\-rois\fR
.IP
<surface> \- the surface to draw on
<limit> \- geodesic distance limit from vertex, in mm
<vertex\-list\-file> \- a text file containing the vertices to draw ROIs
.IP
around
.IP
<metric\-out> \- output \- the output metric
.IP
[\-gaussian] \- generate a gaussian kernel instead of a flat ROI
.IP
<sigma> \- the sigma for the gaussian kernel, in mm
.IP
[\-overlap\-logic] \- how to handle overlapping ROIs, default ALLOW
.IP
<method> \- the method of resolving overlaps
.IP
[\-names] \- name the columns from text file
.IP
<name\-list\-file> \- a text file containing column names, one per line
.IP
For each vertex in the list file, a column in the output metric is
created, and an ROI around that vertex is drawn in that column.  Each
metric column will have zeros outside the geodesic distance spacified by
<limit>, and by default will have a value of 1.0 inside it.  If the
\fB\-gaussian\fR option is specified, the values inside the ROI will instead
form a gaussian with the specified value of sigma, normalized so that the
sum of the nonzero values in the metric column is 1.0.  The <method>
argument to \fB\-overlap\-logic\fR must be one of ALLOW, CLOSEST, or EXCLUDE.
ALLOW is the default, and means that ROIs are treated independently and
may overlap.  CLOSEST means that ROIs may not overlap, and that no ROI
contains vertices that are closer to a different seed vertex.  EXCLUDE
means that ROIs may not overlap, and that any vertex within range of more
than one ROI does not belong to any ROI.
.PP
\fB\-surface\-inflation\fR
SURFACE INFLATION
.IP
wb_command \fB\-surface\-inflation\fR
.IP
<anatomical\-surface\-in> \- the anatomical surface
<surface\-in> \- the surface file to inflate
<number\-of\-smoothing\-cycles> \- number of smoothing cycles
<smoothing\-strength> \- smoothing strength (ranges [0.0 \- 1.0])
<smoothing\-iterations> \- smoothing iterations
<inflation\-factor> \- inflation factor
<surface\-out> \- output \- output surface file
.IP
Inflate a surface by performing cycles that consist of smoothing
followed by inflation (to correct shrinkage caused by smoothing).
.PP
\fB\-surface\-information\fR
DISPLAY INFORMATION ABOUT A SURFACE
.IP
wb_command \fB\-surface\-information\fR
.IP
<Surface File> \- Surface for which information is displayed
.IP
Information about surface is displayed including vertices,
triangles, bounding box, and spacing.
.PP
\fB\-surface\-match\fR
SURFACE MATCH
.IP
wb_command \fB\-surface\-match\fR
.IP
<Match Surface File> \- Match (Reference) Surface
<Input Surface File> \- File containing surface that will be transformed
<Output Surface Name> \- Surface File after transformation
.IP
The Input Surface File will be transformed so that its coordinate ranges
(bounding box) match that of the Match Surface File
.PP
\fB\-surface\-modify\-sphere\fR
CHANGE RADIUS AND OPTIONALLY RECENTER A SPHERE
.IP
wb_command \fB\-surface\-modify\-sphere\fR
.IP
<sphere\-in> \- the sphere to modify
<radius> \- the radius the output sphere should have
<sphere\-out> \- output \- the output sphere
.IP
[\-recenter] \- recenter the sphere by means of the bounding box
.IP
This command may be useful if you have used \fB\-surface\-resample\fR to resample
a sphere, which can suffer from problems generally not present in
\fB\-surface\-sphere\-project\-unproject\fR.  If the sphere should already be
centered around the origin, using \fB\-recenter\fR may still shift it slightly
before changing the radius, which is likely to be undesireable.
.IP
If <sphere\-in> is not close to spherical, or not centered around the
origin and \fB\-recenter\fR is not used, a warning is printed.
.PP
\fB\-surface\-normals\fR
OUTPUT VERTEX NORMALS AS METRIC FILE
.IP
wb_command \fB\-surface\-normals\fR
.IP
<surface> \- the surface to output the normals of
<metric\-out> \- output \- the normal vectors
.IP
Computes the normal vectors of the surface file, and outputs them as a 3
column metric file.
.PP
\fB\-surface\-resample\fR
RESAMPLE A SURFACE TO A DIFFERENT MESH
.IP
wb_command \fB\-surface\-resample\fR
.IP
<surface\-in> \- the surface file to resample
<current\-sphere> \- a sphere surface with the mesh that the input surface
.IP
is currently on
.IP
<new\-sphere> \- a sphere surface that is in register with <current\-sphere>
.IP
and has the desired output mesh
.IP
<method> \- the method name
<surface\-out> \- output \- the output surface file
.IP
[\-area\-surfs] \- specify surfaces to do vertex area correction based on
.IP
<current\-area> \- a relevant surface with <current\-sphere> mesh
<new\-area> \- a relevant surface with <new\-sphere> mesh
.IP
[\-area\-metrics] \- specify vertex area metrics to do area correction based
.IP
on
<current\-area> \- a metric file with vertex areas for <current\-sphere>
.IP
mesh
.IP
<new\-area> \- a metric file with vertex areas for <new\-sphere> mesh
.IP
Resamples a surface file, given two spherical surfaces that are in
register.  If ADAP_BARY_AREA is used, exactly one of \fB\-area\-surfs\fR or
\fB\-area\-metrics\fR must be specified.  This method is not generally
recommended for surface resampling, but is provided for completeness.
.IP
The BARYCENTRIC method is generally recommended for anatomical surfaces,
in order to minimize smoothing.
.IP
For cut surfaces (including flatmaps), use \fB\-surface\-cut\-resample\fR.
.IP
Instead of resampling a spherical surface, the
\fB\-surface\-sphere\-project\-unproject\fR command is recommended.
.IP
The <method> argument must be one of the following:
.IP
ADAP_BARY_AREA
BARYCENTRIC
.PP
\fB\-surface\-smoothing\fR
SURFACE SMOOTHING
.IP
wb_command \fB\-surface\-smoothing\fR
.IP
<surface\-in> \- the surface file to smooth
<smoothing\-strength> \- smoothing strength (ranges [0.0 \- 1.0])
<smoothing\-iterations> \- smoothing iterations
<surface\-out> \- output \- output surface file
.IP
Smooths a surface by averaging vertex coordinates with those of the
neighboring vertices.
.PP
\fB\-surface\-sphere\-project\-unproject\fR
DEFORM A SPHERE ACCORDING TO A REGISTRATION
.IP
wb_command \fB\-surface\-sphere\-project\-unproject\fR
.IP
<sphere\-in> \- the sphere with the desired output mesh
<sphere\-project\-to> \- a sphere that aligns with sphere\-in
<sphere\-unproject\-from> \- sphere\-project\-to deformed to the output space
<sphere\-out> \- output \- the output sphere
.IP
Each vertex of <sphere\-in> is projected to <sphere\-project\-to> to obtain
barycentric weights, which are then used to unproject from
<sphere\-unproject\-from>.  This results in a sphere with the topology of
<sphere\-in>, but coordinates shifted by the deformation between
<sphere\-project\-to> and <sphere\-unproject\-from>.  <sphere\-project\-to> and
<sphere\-unproject\-from> must have the same topology as each other, but
<sphere\-in> may have different topology.
.PP
\fB\-surface\-to\-surface\-3d\-distance\fR
COMPUTE DISTANCE BETWEEN CORRESPONDING VERTICES
.IP
wb_command \fB\-surface\-to\-surface\-3d\-distance\fR
.IP
<surface\-comp> \- the surface to compare to the reference
<surface\-ref> \- the surface to use as the reference
<dists\-out> \- output \- the output distances
.IP
[\-vectors] \- output the displacement vectors
.IP
<vectors\-out> \- output \- the output vectors
.IP
Computes the vector difference between the vertices of each surface with
the same index, as (comp \- ref), and output the magnitudes, and
optionally the displacement vectors.
.PP
\fB\-surface\-vertex\-areas\fR
MEASURE SURFACE AREA EACH VERTEX IS RESPONSIBLE FOR
.IP
wb_command \fB\-surface\-vertex\-areas\fR
.IP
<surface> \- the surface to measure
<metric> \- output \- the output metric
.IP
Each vertex gets one third of the area of each triangle it is a part of.
Units are mm^2.
.PP
\fB\-surface\-wedge\-volume\fR
MEASURE PER\-VERTEX VOLUME BETWEEN SURFACES
.IP
wb_command \fB\-surface\-wedge\-volume\fR
.IP
<inner\-surface> \- the inner surface
<outer\-surface> \- the outer surface
<metric> \- output \- the output metric
.IP
Compute the volume of each vertex's area from one surface to another.
The surfaces must have vertex correspondence.
.PP
\fB\-unit\-test\fR
.PP
\fB\-volume\-affine\-resample\fR
RESAMPLE VOLUME USING AFFINE TRANSFORM
.IP
wb_command \fB\-volume\-affine\-resample\fR
.IP
<volume\-in> \- volume to resample
<affine> \- the affine file to apply
<volume\-space> \- a volume file in the volume space you want for the
.IP
output
.IP
<method> \- the resampling method
<volume\-out> \- output \- the output volume
.IP
[\-flirt] \- MUST be used if affine is a flirt affine
.IP
<source\-volume> \- the source volume used when generating the affine
<target\-volume> \- the target volume used when generating the affine
.TP
Resample a volume file with an affine transformation.
The recommended
.IP
methods are CUBIC (cubic spline) for most data, and ENCLOSING_VOXEL for
label data.  The parameter <method> must be one of:
.IP
CUBIC
ENCLOSING_VOXEL
TRILINEAR
.PP
\fB\-volume\-all\-labels\-to\-rois\fR
MAKE ROIS FROM ALL LABELS IN A VOLUME FRAME
.IP
wb_command \fB\-volume\-all\-labels\-to\-rois\fR
.IP
<label\-in> \- the input volume label file
<map> \- the number or name of the label map to use
<volume\-out> \- output \- the output volume file
.IP
The output volume has a frame for each label in the specified input
frame, other than the ??? label, each of which contains an ROI of all
voxels that are set to the corresponding label.
.PP
\fB\-volume\-capture\-plane\fR
INTERPOLATE IMAGE FROM PLANE THROUGH VOLUME
.IP
wb_command \fB\-volume\-capture\-plane\fR
.IP
<volume> \- the volume file to interpolate from
<subvolume> \- the name or number of the subvolume to use
<interp> \- interpolation type
<h\-dim> \- width of output image, in pixels
<v\-dim> \- height of output image, in pixels
<scale\-min> \- value to render as black
<scale\-max> \- value to render as white
<bottom\-left\-x> \- x\-coordinate of the bottom left of the output image
<bottom\-left\-y> \- y\-coordinate of the bottom left of the output image
<bottom\-left\-z> \- z\-coordinate of the bottom left of the output image
<bottom\-right\-x> \- x\-coordinate of the bottom right of the output image
<bottom\-right\-y> \- y\-coordinate of the bottom right of the output image
<bottom\-right\-z> \- z\-coordinate of the bottom right of the output image
<top\-left\-x> \- x\-coordinate of the top left of the output image
<top\-left\-y> \- y\-coordinate of the top left of the output image
<top\-left\-z> \- z\-coordinate of the top left of the output image
<image> \- output \- the output image
.IP
NOTE: If you want to generate an image with all of the capabilities of
the GUI rendering, see \fB\-show\-scene\fR.
.IP
Renders an image of an arbitrary plane through the volume file, with a
simple linear grayscale palette.  The parameter <interp> must be one of:
.IP
CUBIC
ENCLOSING_VOXEL
TRILINEAR
.PP
\fB\-volume\-copy\-extensions\fR
COPY EXTENDED DATA TO ANOTHER VOLUME FILE
.IP
wb_command \fB\-volume\-copy\-extensions\fR
.IP
<data\-volume> \- the volume file containing the voxel data to use
<extension\-volume> \- the volume file containing the extensions to use
<volume\-out> \- output \- the output volume
.IP
[\-drop\-unknown] \- don't copy extensions that workbench doesn't understand
.IP
This command copies the information in a volume file that isn't a
critical part of the standard header or data matrix, e.g. map names,
palette settings, label tables.  If \fB\-drop\-unknown\fR is not specified, it
also copies similar kinds of information set by other software.
.PP
\fB\-volume\-create\fR
CREATE A BLANK VOLUME FILE
.IP
wb_command \fB\-volume\-create\fR
.IP
<i\-dim> \- length of first dimension
<j\-dim> \- length of second dimension
<k\-dim> \- length of third dimension
<volume\-out> \- output \- the output volume
.IP
[\-plumb] \- set via axis order and spacing/offset
.IP
<axis\-order> \- a string like 'XYZ' that specifies which index is along
.IP
which spatial dimension
.IP
<x\-spacing> \- change in x\-coordinate from incrementing the relevant
.IP
index
.IP
<y\-spacing> \- change in y\-coordinate from incrementing the relevant
.IP
index
.IP
<z\-spacing> \- change in z\-coordinate from incrementing the relevant
.IP
index
.IP
<x\-offset> \- the x\-coordinate of the first voxel
<y\-offset> \- the y\-coordinate of the first voxel
<z\-offset> \- the z\-coordinate of the first voxel
.IP
[\-sform] \- set via a nifti sform
.IP
<xi\-spacing> \- increase in x coordinate from incrementing the i index
<xj\-spacing> \- increase in x coordinate from incrementing the j index
<xk\-spacing> \- increase in x coordinate from incrementing the k index
<x\-offset> \- x coordinate of first voxel
<yi\-spacing> \- increase in y coordinate from incrementing the i index
<yj\-spacing> \- increase in y coordinate from incrementing the j index
<yk\-spacing> \- increase in y coordinate from incrementing the k index
<y\-offset> \- y coordinate of first voxel
<zi\-spacing> \- increase in z coordinate from incrementing the i index
<zj\-spacing> \- increase in z coordinate from incrementing the j index
<zk\-spacing> \- increase in z coordinate from incrementing the k index
<z\-offset> \- z coordinate of first voxel
.TP
Creates a volume file full of zeros.
Exactly one of \fB\-plumb\fR or \fB\-sform\fR
.IP
must be specified.
.PP
\fB\-volume\-dilate\fR
DILATE A VOLUME FILE
.IP
wb_command \fB\-volume\-dilate\fR
.IP
<volume> \- the volume to dilate
<distance> \- distance in mm to dilate
<method> \- dilation method to use
<volume\-out> \- output \- the output volume
.IP
[\-exponent] \- use a different exponent in the weighting function
.IP
<exponent> \- exponent 'n' to use in (1 / (distance ^ n)) as the
.IP
weighting function (default 2)
.IP
[\-bad\-voxel\-roi] \- specify an roi of voxels to overwrite, rather than
.IP
voxels with value zero
<roi\-volume> \- volume file, positive values denote voxels to have
.IP
their values replaced
.IP
[\-data\-roi] \- specify an roi of where there is data
.IP
<roi\-volume> \- volume file, positive values denote voxels that have
.IP
data
.IP
[\-subvolume] \- select a single subvolume to dilate
.IP
<subvol> \- the subvolume number or name
.IP
For all voxels that are designated as bad, if they neighbor a non\-bad
voxel with data or are within the specified distance of such a voxel,
replace the value in the bad voxel with a value calculated from nearby
non\-bad voxels that have data, otherwise set the value to zero.  No
matter how small <distance> is, dilation will always use at least the
face neighbor voxels.
.IP
By default, voxels that have data with the value 0 are bad, specify
\fB\-bad\-voxel\-roi\fR to only count voxels as bad if they are selected by the
roi.  If \fB\-data\-roi\fR is not specified, all voxels are assumed to have data.
.IP
Valid values for <method> are:
.IP
NEAREST \- use the value from the nearest good voxel
WEIGHTED \- use a weighted average based on distance
.PP
\fB\-volume\-erode\fR
ERODE A VOLUME FILE
.IP
wb_command \fB\-volume\-erode\fR
.IP
<volume> \- the volume to erode
<distance> \- distance in mm to erode
<volume\-out> \- output \- the output volume
.IP
[\-roi] \- assume voxels outside this roi are nonzero
.IP
<roi\-volume> \- volume file, positive values denote voxels that have
.IP
data
.IP
[\-subvolume] \- select a single subvolume to dilate
.IP
<subvol> \- the subvolume number or name
.IP
Around each voxel with a value of zero, set surrounding voxels to zero.
The surrounding voxels are all face neighbors and all voxels within the
specified distance (center to center).
.PP
\fB\-volume\-estimate\-fwhm\fR
ESTIMATE FWHM SMOOTHNESS OF A VOLUME
.IP
wb_command \fB\-volume\-estimate\-fwhm\fR
.IP
<volume> \- the input volume
.IP
[\-roi] \- use only data within an ROI
.IP
<roivol> \- the volume to use as an ROI
.IP
[\-subvolume] \- select a single subvolume to estimate smoothness of
.IP
<subvol> \- the subvolume number or name
.IP
Estimates the smoothness of the input volume in X, Y, and Z directions
separately, printing the estimates to standard output.  If \fB\-subvolume\fR is
not specified, each subvolume is estimated and displayed separately.
.PP
\fB\-volume\-extrema\fR
FIND EXTREMA IN A VOLUME FILE
.IP
wb_command \fB\-volume\-extrema\fR
.IP
<volume\-in> \- volume file to find the extrema of
<distance> \- the minimum distance between identified extrema of the same
.IP
type
.IP
<volume\-out> \- output \- the output extrema volume
.IP
[\-presmooth] \- smooth the volume before finding extrema
.IP
<kernel> \- the sigma for the gaussian smoothing kernel, in mm
.IP
[\-roi] \- ignore values outside the selected area
.IP
<roi\-volume> \- the area to find extrema in
.IP
[\-threshold] \- ignore small extrema
.IP
<low> \- the largest value to consider for being a minimum
<high> \- the smallest value to consider for being a maximum
.IP
[\-sum\-subvols] \- output the sum of the extrema subvolumes instead of each
.IP
subvolume separately
.IP
[\-consolidate\-mode] \- use consolidation of local minima instead of a
.IP
large neighborhood
.IP
[\-only\-maxima] \- only find the maxima
.IP
[\-only\-minima] \- only find the minima
.IP
[\-subvolume] \- select a single subvolume to find extrema in
.IP
<subvolume> \- the subvolume number or name
.IP
Finds extrema in a volume file, such that no two extrema of the same type
are within <distance> of each other.  The extrema are labeled as \fB\-1\fR for
minima, 1 for maxima, 0 otherwise.  If \fB\-only\-maxima\fR or \fB\-only\-minima\fR is
specified, then it will ignore extrema not of the specified type.  These
options are mutually exclusive.
.IP
If \fB\-sum\-subvols\fR is specified, these extrema subvolumes are summed, and
the output has a single subvolume with this result.
.IP
By default, a datapoint is an extrema only if it is more extreme than
every other datapoint that is within <distance> from it.  If
\fB\-consolidate\-mode\fR is used, it instead starts by finding all datapoints
that are more extreme than their immediate neighbors, then while there
are any extrema within <distance> of each other, take the two extrema
closest to each other and merge them into one by a weighted average based
on how many original extrema have been merged into each.
.IP
By default, all input subvolumes are used with no smoothing, use
\fB\-subvolume\fR to specify a single subvolume to use, and \fB\-presmooth\fR to smooth
the input before finding the extrema.
.PP
\fB\-volume\-fill\-holes\fR
FILL HOLES IN AN ROI VOLUME
.IP
wb_command \fB\-volume\-fill\-holes\fR
.IP
<volume\-in> \- the input ROI volume
<volume\-out> \- output \- the output ROI volume
.IP
Finds all face\-connected parts that are not included in the ROI, and
fills all but the largest one with ones.
.PP
\fB\-volume\-find\-clusters\fR
FILTER CLUSTERS BY VOLUME
.IP
wb_command \fB\-volume\-find\-clusters\fR
.IP
<volume\-in> \- the input volume
<value\-threshold> \- threshold for data values
<minimum\-volume> \- threshold for cluster volume, in mm^3
<volume\-out> \- output \- the output volume
.IP
[\-less\-than] \- find values less than <value\-threshold>, rather than
.IP
greater
.IP
[\-roi] \- select a region of interest
.IP
<roi\-volume> \- the roi, as a volume file
.IP
[\-subvolume] \- select a single subvolume
.IP
<subvol> \- the subvolume number or name
.IP
[\-size\-ratio] \- ignore clusters smaller than a given fraction of the
.IP
largest cluster in map
<ratio> \- fraction of the largest cluster's volume
.IP
[\-distance] \- ignore clusters further than a given distance from the
.IP
largest cluster
<distance> \- how far from the largest cluster a cluster can be, edge
.IP
to edge, in mm
.IP
[\-start] \- start labeling clusters from a value other than 1
.IP
<startval> \- the value to give the first cluster found
.IP
Outputs a volume with nonzero integers for all voxels within a large
enough cluster, and zeros elsewhere.  The integers denote cluster
membership (by default, first cluster found will use value 1, second
cluster 2, etc).  By default, values greater than <value\-threshold> are
considered to be in a cluster, use \fB\-less\-than\fR to test for values less
than the threshold.  To apply this as a mask to the data, or to do more
complicated thresholding, see \fB\-volume\-math\fR.
.PP
\fB\-volume\-gradient\fR
GRADIENT OF A VOLUME FILE
.IP
wb_command \fB\-volume\-gradient\fR
.IP
<volume\-in> \- the input volume
<volume\-out> \- output \- the output gradient magnitude volume
.IP
[\-presmooth] \- smooth the volume before computing the gradient
.IP
<kernel> \- sigma for gaussian weighting function, in mm
.IP
[\-roi] \- select a region of interest to take the gradient of
.IP
<roi\-volume> \- the region to take the gradient within
.IP
[\-vectors] \- output vectors
.IP
<vector\-volume\-out> \- output \- the vectors as a volume file
.IP
[\-subvolume] \- select a single subvolume to take the gradient of
.IP
<subvol> \- the subvolume number or name
.IP
Computes the gradient of the volume by doing linear regressions for each
voxel, considering only its face neighbors unless too few face neighbors
exist.  The gradient vector is constructed from the partial derivatives
of the resulting linear function, and the magnitude of this vector is the
output.  If specified, the volume vector output is arranged with the x,
y, and z components from a subvolume as consecutive subvolumes.
.PP
\fB\-volume\-label\-export\-table\fR
EXPORT LABEL TABLE FROM VOLUME AS TEXT
.IP
wb_command \fB\-volume\-label\-export\-table\fR
.IP
<label\-in> \- the input volume label file
<map> \- the number or name of the label map to use
<table\-out> \- output \- the output text file
.IP
Takes the label table from the volume label map, and writes it to a text
format matching what is expected by \fB\-volume\-label\-import\fR.
.PP
\fB\-volume\-label\-import\fR
IMPORT A LABEL VOLUME TO CARET FORMAT
.IP
wb_command \fB\-volume\-label\-import\fR
.IP
<input> \- the label volume to import
<label\-list\-file> \- text file containing the values and names for labels
<output> \- output \- the output workbench label volume
.IP
[\-discard\-others] \- set any voxels with values not mentioned in the label
.IP
list to the ??? label
.IP
[\-unlabeled\-value] \- set the value that will be interpreted as unlabeled
.IP
<value> \- the numeric value for unlabeled (default 0)
.IP
[\-subvolume] \- select a single subvolume to import
.IP
<subvol> \- the subvolume number or name
.IP
[\-drop\-unused\-labels] \- remove any unused label values from the label
.IP
table
.IP
Creates a new volume with label information in the header in the caret
nifti extension format.  You may specify the empty string ('' will work
on linux/mac) for <label\-list\-file>, which will be treated as if it is an
empty file.  The label list file must have lines of the following format:
.IP
<labelname>
<value> <red> <green> <blue> <alpha>
.IP
Do not specify the "unlabeled" key in the file, it is assumed that 0
means not labeled unless \fB\-unlabeled\-value\fR is specified.  Label names must
be on a separate line, but may contain spaces or other unusual characters
(but not newline).  Whitespace is trimmed from both ends of the label
name, but is kept if it is in the middle of a label.  The values of red,
green, blue and alpha must be integers from 0 to 255, and will specify
the color the label is drawn as (alpha of 255 means opaque, which is
probably what you want).  By default, it will set new label names with
names of LABEL_# for any values encountered that are not mentioned in the
list file, specify \fB\-discard\-others\fR to instead set these voxels to the
"unlabeled" key.
.PP
\fB\-volume\-label\-to\-roi\fR
MAKE A VOLUME LABEL INTO AN ROI VOLUME
.IP
wb_command \fB\-volume\-label\-to\-roi\fR
.IP
<label\-in> \- the input volume label file
<volume\-out> \- output \- the output volume file
.IP
[\-name] \- select label by name
.IP
<label\-name> \- the label name that you want an roi of
.IP
[\-key] \- select label by key
.IP
<label\-key> \- the label key that you want an roi of
.IP
[\-map] \- select a single label map to use
.IP
<map> \- the map number or name
.IP
For each map in <label\-in>, a map is created in <volume\-out> where all
locations labeled with <label\-name> or with a key of <label\-key> are
given a value of 1, and all other locations are given 0.  Exactly one of
\fB\-name\fR and \fB\-key\fR must be specified.  Specify \fB\-map\fR to use only one map from
<label\-in>.
.PP
\fB\-volume\-label\-to\-surface\-mapping\fR
MAP A LABEL VOLUME TO A SURFACE LABEL FILE
.IP
wb_command \fB\-volume\-label\-to\-surface\-mapping\fR
.IP
<volume> \- the volume to map data from
<surface> \- the surface to map the data onto
<label\-out> \- output \- the output gifti label file
.IP
[\-ribbon\-constrained] \- use ribbon constrained mapping algorithm
.IP
<inner\-surf> \- the inner surface of the ribbon
<outer\-surf> \- the outer surface of the ribbon
.IP
[\-volume\-roi] \- use a volume roi
.IP
<roi\-volume> \- the volume file
.IP
[\-voxel\-subdiv] \- voxel divisions while estimating voxel weights
.IP
<subdiv\-num> \- number of subdivisions, default 3
.IP
[\-subvol\-select] \- select a single subvolume to map
.IP
<subvol> \- the subvolume number or name
.TP
Map label volume data to a surface.
If \fB\-ribbon\-constrained\fR is not
.IP
specified, uses the enclosing voxel method.
.PP
\fB\-volume\-math\fR
EVALUATE EXPRESSION ON VOLUME FILES
.IP
wb_command \fB\-volume\-math\fR
.IP
<expression> \- the expression to evaluate, in quotes
<volume\-out> \- output \- the output volume
.IP
[\-fixnan] \- replace NaN results with a value
.IP
<replace> \- value to replace NaN with
.IP
[\-var] \- repeatable \- a volume file to use as a variable
.IP
<name> \- the name of the variable, as used in the expression
<volume> \- the volume file to use as this variable
.IP
[\-subvolume] \- select a single subvolume
.IP
<subvol> \- the subvolume number or name
.IP
[\-repeat] \- reuse a single subvolume for each subvolume of calculation
.TP
This command evaluates <expression> at each voxel independently.
There
.IP
must be at least one \fB\-var\fR option (to get the volume space from), even if
the <name> specified in it isn't used in <expression>.  All volumes must
have the same volume space.  Filenames are not valid in <expression>, use
a variable name and a \fB\-var\fR option with matching <name> to specify an
input file.  If the \fB\-subvolume\fR option is given to any \fB\-var\fR option, only
one subvolume is used from that file.  If \fB\-repeat\fR is specified, the file
must either have only one subvolume, or have the \fB\-subvolume\fR option
specified.  All files that don't use \fB\-repeat\fR must have the same number of
subvolumes requested to be used.  The format of <expression> is as
follows:
.IP
Expressions consist of constants, variables, operators, parentheses, and
functions, in infix notation, such as 'exp(\fB\-x\fR + 3) * scale'.  Variables
are strings of any length, using the characters a\-z, A\-Z, 0\-9, and _, but
may not take the name of a named constant.  Currently, there is only one
named constant, PI.  The operators are +, \-, *, /, ^, >, <, >=, <=, ==,
!=, !, &&, ||.  These behave as in C, except that ^ is exponentiation,
i.e. pow(x, y), and takes higher precedence than other binary operators
(also, '\-3^\-4^\-5' means '\-(3^(\-(4^\-5)))').  The <=, >=, ==, and !=
operators are given a small amount of wiggle room, equal to one millionth
of the smaller of the absolute values of the values being compared.
.IP
Comparison and logical operators return 0 or 1, you can do masking with
expressions like 'x * (mask > 0)'.  For all logical operators, an input
is considered true iff it is greater than 0.  The expression '0 < x < 5'
is not syntactically wrong, but it will NOT do what is desired, because
it is evaluated left to right, i.e. '((0 < x) < 5)', which will always
return 1, as both possible results of a comparison are less than 5.  A
warning is generated if an expression of this type is detected.  Use
something like 'x > 0 && x < 5' to get the desired behavior.
.IP
Whitespace between elements is ignored, ' sin ( 2 * x ) ' is equivalent
to 'sin(2*x)', but 's in(2*x)' is an error.  Implied multiplication is
not allowed, the expression '2x' will be parsed as a variable.
Parentheses are (), do not use [] or {}.  Functions require parentheses,
the expression 'sin x' is an error.
.IP
The following functions are supported:
.IP
sin: 1 argument, the sine of the argument (units are radians)
cos: 1 argument, the cosine of the argument (units are radians)
tan: 1 argument, the tangent of the argument (units are radians)
asin: 1 argument, the inverse of sine of the argument, in radians
acos: 1 argument, the inverse of cosine of the argument, in radians
atan: 1 argument, the inverse of tangent of the argument, in radians
atan2: 2 arguments, atan2(y, x) returns the inverse of tangent of
.IP
(y/x), in radians, determining quadrant by the sign of both arguments
.IP
sinh: 1 argument, the hyperbolic sine of the argument
cosh: 1 argument, the hyperbolic cosine of the argument
tanh: 1 argument, the hyperboloc tangent of the argument
asinh: 1 argument, the inverse hyperbolic sine of the argument
acosh: 1 argument, the inverse hyperbolic cosine of the argument
atanh: 1 argument, the inverse hyperboloc tangent of the argument
ln: 1 argument, the natural logarithm of the argument
exp: 1 argument, the constant e raised to the power of the argument
log: 1 argument, the base 10 logarithm of the argument
sqrt: 1 argument, the square root of the argument
abs: 1 argument, the absolute value of the argument
floor: 1 argument, the largest integer not greater than the argument
round: 1 argument, the nearest integer, with ties rounded away from
.IP
zero
.IP
ceil: 1 argument, the smallest integer not less than the argument
min: 2 arguments, min(x, y) returns y if (x > y), x otherwise
max: 2 arguments, max(x, y) returns y if (x < y), x otherwise
mod: 2 arguments, mod(x, y) = x \- y * floor(x / y), or 0 if y == 0
clamp: 3 arguments, clamp(x, low, high) = min(max(x, low), high)
.PP
\fB\-volume\-merge\fR
MERGE VOLUME FILES INTO A NEW FILE
.IP
wb_command \fB\-volume\-merge\fR
.IP
<volume\-out> \- output \- the output volume file
.IP
[\-volume] \- repeatable \- specify an input volume file
.IP
<volume\-in> \- a volume file to use subvolumes from
.IP
[\-subvolume] \- repeatable \- select a single subvolume to use
.IP
<subvol> \- the subvolume number or name
.IP
[\-up\-to] \- use an inclusive range of subvolumes
.IP
<last\-subvol> \- the number or name of the last subvolume to
.IP
include
.IP
[\-reverse] \- use the range in reverse order
.IP
Takes one or more volume files and constructs a new volume file by
concatenating subvolumes from them.  The input volume files must have the
same volume space.
.IP
Example: wb_command \fB\-volume\-merge\fR out.nii \fB\-volume\fR first.nii \fB\-subvolume\fR 1
\fB\-volume\fR second.nii
.IP
This example would take the first subvolume from first.nii, followed by
all subvolumes from second.nii, and write these to out.nii.
.PP
\fB\-volume\-palette\fR
SET THE PALETTE OF A VOLUME FILE
.IP
wb_command \fB\-volume\-palette\fR
.IP
<volume> \- the volume file to modify
<mode> \- the mapping mode
.IP
[\-subvolume] \- select a single subvolume
.IP
<subvolume> \- the subvolume number or name
.IP
[\-pos\-percent] \- percentage min/max for positive data coloring
.IP
<pos\-min\-%> \- the percentile for the least positive data
<pos\-max\-%> \- the percentile for the most positive data
.IP
[\-neg\-percent] \- percentage min/max for negative data coloring
.IP
<neg\-min\-%> \- the percentile for the least negative data
<neg\-max\-%> \- the percentile for the most negative data
.IP
[\-pos\-user] \- user min/max values for positive data coloring
.IP
<pos\-min\-user> \- the value for the least positive data
<pos\-max\-user> \- the value for the most positive data
.IP
[\-neg\-user] \- user min/max values for negative data coloring
.IP
<neg\-min\-user> \- the value for the least negative data
<neg\-max\-user> \- the value for the most negative data
.IP
[\-interpolate] \- interpolate colors
.IP
<interpolate> \- boolean, whether to interpolate
.IP
[\-disp\-pos] \- display positive data
.IP
<display> \- boolean, whether to display
.IP
[\-disp\-neg] \- display positive data
.IP
<display> \- boolean, whether to display
.IP
[\-disp\-zero] \- display data closer to zero than the min cutoff
.IP
<display> \- boolean, whether to display
.IP
[\-palette\-name] \- set the palette used
.IP
<name> \- the name of the palette
.IP
[\-thresholding] \- set the thresholding
.IP
<type> \- thresholding setting
<test> \- show values inside or outside thresholds
<min> \- lower threshold
<max> \- upper threshold
.TP
The original volume file is overwritten with the modified version.
By
.IP
default, all columns of the volume file are adjusted to the new settings,
use the \fB\-subvolume\fR option to change only one subvolume.  Mapping settings
not specified in options will be taken from the first subvolume.  The
<mode> argument must be one of the following:
.IP
MODE_AUTO_SCALE
MODE_AUTO_SCALE_ABSOLUTE_PERCENTAGE
MODE_AUTO_SCALE_PERCENTAGE
MODE_USER_SCALE
.IP
The <name> argument to \fB\-palette\-name\fR must be one of the following:
.IP
PSYCH
PSYCH\-NO\-NONE
ROY\-BIG
ROY\-BIG\-BL
Orange\-Yellow
Gray_Interp_Positive
Gray_Interp
clear_brain
videen_style
fidl
raich4_clrmid
raich6_clrmid
HSB8_clrmid
RBGYR20
RBGYR20P
POS_NEG
red\-yellow
blue\-lightblue
FSL
power_surf
fsl_red
fsl_green
fsl_blue
fsl_yellow
JET256
.IP
The <type> argument to \fB\-thresholding\fR must be one of the following:
.IP
THRESHOLD_TYPE_OFF
THRESHOLD_TYPE_NORMAL
.IP
The <test> argument to \fB\-thresholding\fR must be one of the following:
.IP
THRESHOLD_TEST_SHOW_OUTSIDE
THRESHOLD_TEST_SHOW_INSIDE
.PP
\fB\-volume\-parcel\-resampling\fR
SMOOTH AND RESAMPLE VOLUME PARCELS
.IP
wb_command \fB\-volume\-parcel\-resampling\fR
.IP
<volume\-in> \- the input data volume
<cur\-parcels> \- label volume of where the parcels currently are
<new\-parcels> \- label volume of where the parcels should be
<kernel> \- gaussian kernel sigma to smooth by during resampling
<volume\-out> \- output \- output volume
.IP
[\-fix\-zeros] \- treat zero values as not being data
.IP
[\-subvolume] \- select a single subvolume as input
.IP
<subvol> \- the subvolume number or name
.IP
Smooths and resamples the region inside each label in cur\-parcels to the
region of the same label name in new\-parcels.  Any voxels in the output
label region but outside the input label region will be extrapolated from
nearby data.  The \fB\-fix\-zeros\fR option causes the smoothing to not use an
input value if it is zero, but still write a smoothed value to the voxel,
and after smoothing is complete, it will check for any remaining values
of zero, and fill them in with extrapolated values.
.TP
Note: all volumes must have the same dimensions and spacing.
To use a
.IP
different output space, see \fB\-volume\-parcel\-resampling\-generic\fR.
.PP
\fB\-volume\-parcel\-resampling\-generic\fR
SMOOTH AND RESAMPLE VOLUME PARCELS FROM DIFFERENT VOLUME SPACE
.IP
wb_command \fB\-volume\-parcel\-resampling\-generic\fR
.IP
<volume\-in> \- the input data volume
<cur\-parcels> \- label volume of where the parcels currently are
<new\-parcels> \- label volume of where the parcels should be
<kernel> \- gaussian kernel sigma to smooth by during resampling
<volume\-out> \- output \- output volume
.IP
[\-fix\-zeros] \- treat zero values as not being data
.IP
[\-subvolume] \- select a single subvolume as input
.IP
<subvol> \- the subvolume number or name
.IP
Smooths and resamples the region inside each label in cur\-parcels to the
region of the same label name in new\-parcels.  Any voxels in the output
label region but outside the input label region will be extrapolated from
nearby data.  The \fB\-fix\-zeros\fR option causes the smoothing to not use an
input value if it is zero, but still write a smoothed value to the voxel,
and after smoothing is complete, it will check for any remaining values
of zero, and fill them in with extrapolated values.  The output volume
will use the volume space of new\-parcels, which does not need to be in
the same volume space as the input.
.PP
\fB\-volume\-parcel\-smoothing\fR
SMOOTH PARCELS IN A VOLUME SEPARATELY
.IP
wb_command \fB\-volume\-parcel\-smoothing\fR
.IP
<data\-volume> \- the volume to smooth
<label\-volume> \- a label volume containing the parcels to smooth
<kernel> \- the gaussian smoothing kernel sigma, in mm
<volume\-out> \- output \- the output volume
.IP
[\-fix\-zeros] \- treat zero values as not being data
.IP
[\-subvolume] \- select a single subvolume to smooth
.IP
<subvol> \- the subvolume number or name
.IP
The volume is smoothed within each label in the label volume using data
only from within the label.  Equivalent to running volume smoothing with
ROIs matching each label separately, then adding the resulting volumes,
but faster.
.PP
\fB\-volume\-reduce\fR
PERFORM REDUCTION OPERATION ACROSS SUBVOLUMES
.IP
wb_command \fB\-volume\-reduce\fR
.IP
<volume\-in> \- the volume file to reduce
<operation> \- the reduction operator to use
<volume\-out> \- output \- the output volume
.IP
[\-exclude\-outliers] \- exclude non\-numeric values and outliers by standard
.IP
deviation
<sigma\-below> \- number of standard deviations below the mean to
.IP
include
.IP
<sigma\-above> \- number of standard deviations above the mean to
.IP
include
.IP
[\-only\-numeric] \- exclude non\-numeric values
.IP
For each voxel, takes the data across subvolumes as a vector, and
performs the specified reduction on it, putting the result into the
single output volume at that voxel.  The reduction operators are as
follows:
.IP
MAX: the maximum value
MIN: the minimum value
INDEXMAX: the 1\-based index of the maximum value
INDEXMIN: the 1\-based index of the minimum value
SUM: add all values
PRODUCT: multiply all values
MEAN: the mean of the data
STDEV: the standard deviation (N denominator)
SAMPSTDEV: the sample standard deviation (N\-1 denominator)
VARIANCE: the variance of the data
TSNR: mean divided by sample standard deviation (N\-1 denominator)
COV: sample standard deviation (N\-1 denominator) divided by mean
MEDIAN: the median of the data
MODE: the mode of the data
COUNT_NONZERO: the number of nonzero elements in the data
.PP
\fB\-volume\-remove\-islands\fR
REMOVE ISLANDS FROM AN ROI VOLUME
.IP
wb_command \fB\-volume\-remove\-islands\fR
.IP
<volume\-in> \- the input ROI volume
<volume\-out> \- output \- the output ROI volume
.IP
Finds all face\-connected parts of the ROI, and zeros out all but the
largest one.
.PP
\fB\-volume\-reorient\fR
CHANGE VOXEL ORDER OF A VOLUME FILE
.IP
wb_command \fB\-volume\-reorient\fR
.IP
<volume> \- the volume to reorient
<orient\-string> \- the desired orientation
<volume\-out> \- out \- the reoriented volume
.IP
Changes the voxel order and the header spacing/origin information such
that the value of any spatial point is unchanged.  Orientation strings
look like 'LPI', which means first index is left to right, second is
posterior to anterior, and third is inferior to superior.  The valid
characters are:
.TP
L
left to right
.TP
R
right to left
.TP
P
posterior to anterior
.TP
A
anterior to posterior
.TP
I
inferior to superior
.TP
S
superior to inferior
.PP
\fB\-volume\-rois\-from\-extrema\fR
CREATE VOLUME ROI MAPS FROM EXTREMA MAPS
.IP
wb_command \fB\-volume\-rois\-from\-extrema\fR
.IP
<volume\-in> \- the input volume
<limit> \- distance limit from voxel center, in mm
<volume\-out> \- output \- the output volume
.IP
[\-gaussian] \- generate a gaussian kernel instead of a flat ROI
.IP
<sigma> \- the sigma for the gaussian kernel, in mm
.IP
[\-roi] \- select a region of interest to use
.IP
<roi\-volume> \- the region to use
.IP
[\-overlap\-logic] \- how to handle overlapping ROIs, default ALLOW
.IP
<method> \- the method of resolving overlaps
.IP
[\-subvolume] \- select a single subvolume to take the gradient of
.IP
<subvol> \- the subvolume number or name
.IP
For each nonzero value in each map, make a map with an ROI around that
location.  If the \fB\-gaussian\fR option is specified, then normalized gaussian
kernels are output instead of ROIs.  The <method> argument to
\fB\-overlap\-logic\fR must be one of ALLOW, CLOSEST, or EXCLUDE.  ALLOW is the
default, and means that ROIs are treated independently and may overlap.
CLOSEST means that ROIs may not overlap, and that no ROI contains
vertices that are closer to a different seed vertex.  EXCLUDE means that
ROIs may not overlap, and that any vertex within range of more than one
ROI does not belong to any ROI.
.PP
\fB\-volume\-set\-space\fR
CHANGE VOLUME SPACE INFORMATION
.IP
wb_command \fB\-volume\-set\-space\fR
.IP
<volume\-in> \- the input volume
<volume\-out> \- output \- the output volume
.IP
[\-plumb] \- set via axis order and spacing/offset
.IP
<axis\-order> \- a string like 'XYZ' that specifies which index is along
.IP
which spatial dimension
.IP
<x\-spacing> \- change in x\-coordinate from incrementing the relevant
.IP
index
.IP
<y\-spacing> \- change in y\-coordinate from incrementing the relevant
.IP
index
.IP
<z\-spacing> \- change in z\-coordinate from incrementing the relevant
.IP
index
.IP
<x\-offset> \- the x\-coordinate of the first voxel
<y\-offset> \- the y\-coordinate of the first voxel
<z\-offset> \- the z\-coordinate of the first voxel
.IP
[\-sform] \- set via a nifti sform
.IP
<xi\-spacing> \- increase in x coordinate from incrementing the i index
<xj\-spacing> \- increase in x coordinate from incrementing the j index
<xk\-spacing> \- increase in x coordinate from incrementing the k index
<x\-offset> \- x coordinate of first voxel
<yi\-spacing> \- increase in y coordinate from incrementing the i index
<yj\-spacing> \- increase in y coordinate from incrementing the j index
<yk\-spacing> \- increase in y coordinate from incrementing the k index
<y\-offset> \- y coordinate of first voxel
<zi\-spacing> \- increase in z coordinate from incrementing the i index
<zj\-spacing> \- increase in z coordinate from incrementing the j index
<zk\-spacing> \- increase in z coordinate from incrementing the k index
<z\-offset> \- z coordinate of first voxel
.IP
Writes a copy of the volume file, with the spacing information changed as
specified.  No reordering of the voxel data occurs.  Exactly one of
\fB\-plumb\fR or \fB\-sform\fR must be specified.
.PP
\fB\-volume\-smoothing\fR
SMOOTH A VOLUME FILE
.IP
wb_command \fB\-volume\-smoothing\fR
.IP
<volume\-in> \- the volume to smooth
<kernel> \- the gaussian smoothing kernel sigma, in mm
<volume\-out> \- output \- the output volume
.IP
[\-roi] \- smooth only from data within an ROI
.IP
<roivol> \- the volume to use as an ROI
.IP
[\-fix\-zeros] \- treat zero values as not being data
.IP
[\-subvolume] \- select a single subvolume to smooth
.IP
<subvol> \- the subvolume number or name
.TP
Gaussian smoothing for volumes.
By default, smooths all subvolumes with
.IP
no ROI, if ROI is given, only positive voxels in the ROI volume have
their values used, and all other voxels are set to zero.  Smoothing a
non\-orthogonal volume will be significantly slower, because the operation
cannot be separated into 1\-dimensional smoothings without distorting the
kernel shape.
.IP
The \fB\-fix\-zeros\fR option causes the smoothing to not use an input value if
it is zero, but still write a smoothed value to the voxel.  This is
useful for zeros that indicate lack of information, preventing them from
pulling down the intensity of nearby voxels, while giving the zero an
extrapolated value.
.PP
\fB\-volume\-stats\fR
SPATIAL STATISTICS ON A VOLUME FILE
.IP
wb_command \fB\-volume\-stats\fR
.IP
<volume\-in> \- the input volume
.IP
[\-reduce] \- use a reduction operation
.IP
<operation> \- the reduction operation
.IP
[\-percentile] \- give the value at a percentile
.IP
<percent> \- the percentile to find
.IP
[\-subvolume] \- only display output for one subvolume
.IP
<subvolume> \- the subvolume number or name
.IP
[\-roi] \- only consider data inside an roi
.IP
<roi\-volume> \- the roi, as a volume file
.IP
[\-match\-maps] \- each subvolume of input uses the corresponding
.IP
subvolume from the roi file
.IP
[\-show\-map\-name] \- print map index and name before each output
.IP
For each subvolume of the input, a single number is printed, resulting
from the specified reduction or percentile operation.  Use \fB\-subvolume\fR to
only give output for a single subvolume.  Use \fB\-roi\fR to consider only the
data within a region.  Exactly one of \fB\-reduce\fR or \fB\-percentile\fR must be
specified.
.IP
The argument to the \fB\-reduce\fR option must be one of the following:
.IP
MAX: the maximum value
MIN: the minimum value
INDEXMAX: the 1\-based index of the maximum value
INDEXMIN: the 1\-based index of the minimum value
SUM: add all values
PRODUCT: multiply all values
MEAN: the mean of the data
STDEV: the standard deviation (N denominator)
SAMPSTDEV: the sample standard deviation (N\-1 denominator)
VARIANCE: the variance of the data
TSNR: mean divided by sample standard deviation (N\-1 denominator)
COV: sample standard deviation (N\-1 denominator) divided by mean
MEDIAN: the median of the data
MODE: the mode of the data
COUNT_NONZERO: the number of nonzero elements in the data
.PP
\fB\-volume\-tfce\fR
DO TFCE ON A VOLUME FILE
.IP
wb_command \fB\-volume\-tfce\fR
.IP
<volume\-in> \- the volume to run TFCE on
<volume\-out> \- output \- the output volume
.IP
[\-presmooth] \- smooth the volume before running TFCE
.IP
<kernel> \- the sigma for the gaussian smoothing kernel, in mm
.IP
[\-roi] \- select a region of interest to run TFCE on
.IP
<roi\-volume> \- the area to run TFCE on, as a volume
.IP
[\-parameters] \- set parameters for TFCE integral
.IP
<E> \- exponent for cluster volume (default 0.5)
<H> \- exponent for threshold value (default 2.0)
.IP
[\-subvolume] \- select a single subvolume
.IP
<subvolume> \- the subvolume number or name
.IP
Threshold\-free cluster enhancement is a method to increase the relative
value of regions that would form clusters in a standard thresholding
test.  This is accomplished by evaluating the integral of:
.IP
e(h, p)^E * h^H * dh
.IP
at each vertex p, where h ranges from 0 to the maximum value in the data,
and e(h, p) is the extent of the cluster containing vertex p at threshold
h.  Negative values are similarly enhanced by negating the data, running
the same process, and negating the result.
.IP
This method is explained in: Smith SM, Nichols TE., "Threshold\-free
cluster enhancement: addressing problems of smoothing, threshold
dependence and localisation in cluster inference." Neuroimage. 2009 Jan
1;44(1):83\-98. PMID: 18501637
.PP
\fB\-volume\-to\-surface\-mapping\fR
MAP VOLUME TO SURFACE
.IP
wb_command \fB\-volume\-to\-surface\-mapping\fR
.IP
<volume> \- the volume to map data from
<surface> \- the surface to map the data onto
<metric\-out> \- output \- the output metric file
.IP
[\-trilinear] \- use trilinear volume interpolation
.IP
[\-enclosing] \- use value of the enclosing voxel
.IP
[\-cubic] \- use cubic splines
.IP
[\-ribbon\-constrained] \- use ribbon constrained mapping algorithm
.IP
<inner\-surf> \- the inner surface of the ribbon
<outer\-surf> \- the outer surface of the ribbon
.IP
[\-volume\-roi] \- use a volume roi
.IP
<roi\-volume> \- the volume file
.IP
[\-voxel\-subdiv] \- voxel divisions while estimating voxel weights
.IP
<subdiv\-num> \- number of subdivisions, default 3
.IP
[\-output\-weights] \- write the voxel weights for a vertex to a volume
.IP
file
<vertex> \- the vertex number to get the voxel weights for, 0\-based
<weights\-out> \- output \- volume to write the weights to
.IP
[\-output\-weights\-text] \- write the voxel weights for all vertices to a
.IP
text file
<text\-out> \- output \- the output text filename
.IP
[\-myelin\-style] \- use the method from myelin mapping
.IP
<ribbon\-roi> \- an roi volume of the cortical ribbon for this
.IP
hemisphere
.IP
<thickness> \- a metric file of cortical thickness
<sigma> \- gaussian kernel in mm for weighting voxels within range
.IP
[\-subvol\-select] \- select a single subvolume to map
.IP
<subvol> \- the subvolume number or name
.TP
You must specify exactly one mapping method.
Enclosing voxel uses the
.IP
value from the voxel the vertex lies inside, while trilinear does a 3D
linear interpolation based on the voxels immediately on each side of the
vertex's position.
.IP
The ribbon mapping method constructs a polyhedron from the vertex's
neighbors on each surface, and estimates the amount of this polyhedron's
volume that falls inside any nearby voxels, to use as the weights for
sampling.  The volume ROI is useful to exclude partial volume effects of
voxels the surfaces pass through, and will cause the mapping to ignore
voxels that don't have a positive value in the mask.  The subdivision
number specifies how it approximates the amount of the volume the
polyhedron intersects, by splitting each voxel into NxNxN pieces, and
checking whether the center of each piece is inside the polyhedron.  If
you have very large voxels, consider increasing this if you get zeros in
your output.
.IP
The myelin style method uses part of the caret5 myelin mapping command to
do the mapping: for each surface vertex, take all voxels closer than the
thickness at the vertex that are within the ribbon ROI, and less than
half the thickness value away from the vertex along the direction of the
surface normal, and apply a gaussian kernel with the specified sigma to
them to get the weights to use.
.PP
\fB\-volume\-vector\-operation\fR
DO A VECTOR OPERATION ON VOLUME FILES
.IP
wb_command \fB\-volume\-vector\-operation\fR
.IP
<vectors\-a> \- first vector input file
<vectors\-b> \- second vector input file
<operation> \- what vector operation to do
<volume\-out> \- output \- the output file
.IP
[\-normalize\-a] \- normalize vectors of first input
.IP
[\-normalize\-b] \- normalize vectors of second input
.IP
[\-normalize\-output] \- normalize output vectors (not valid for dot
.IP
product)
.IP
[\-magnitude] \- output the magnitude of the result (not valid for dot
.IP
product)
.IP
Does a vector operation on two volume files (that must have a multiple of
3 subvolumes).  Either of the inputs may have multiple vectors (more than
3 subvolumes), but not both (at least one must have exactly 3
subvolumes).  The \fB\-magnitude\fR and \fB\-normalize\-output\fR options may not be
specified together, or with the DOT operation.  The <operation> parameter
must be one of the following:
.IP
DOT
CROSS
ADD
SUBTRACT
.PP
\fB\-volume\-warpfield\-resample\fR
RESAMPLE VOLUME USING WARPFIELD
.IP
wb_command \fB\-volume\-warpfield\-resample\fR
.IP
<volume\-in> \- volume to resample
<warpfield> \- the warpfield to apply
<volume\-space> \- a volume file in the volume space you want for the
.IP
output
.IP
<method> \- the resampling method
<volume\-out> \- output \- the output volume
.IP
[\-fnirt] \- MUST be used if using a fnirt warpfield
.IP
<source\-volume> \- the source volume used when generating the warpfield
.TP
Resample a volume file with a warpfield.
The recommended methods are
.IP
CUBIC (cubic spline) for most data, and ENCLOSING_VOXEL for label data.
The parameter <method> must be one of:
.IP
CUBIC
ENCLOSING_VOXEL
TRILINEAR
.PP
\fB\-volume\-weighted\-stats\fR
WEIGHTED SPATIAL STATISTICS ON A VOLUME FILE
.IP
wb_command \fB\-volume\-weighted\-stats\fR
.IP
<volume\-in> \- the input volume
.IP
[\-weight\-volume] \- use weights from a volume file
.IP
<weight\-volume> \- volume file containing the weights
.IP
[\-subvolume] \- only display output for one subvolume
.IP
<subvolume> \- the subvolume number or name
.IP
[\-roi] \- only consider data inside an roi
.IP
<roi\-volume> \- the roi, as a volume file
.IP
[\-match\-maps] \- each subvolume of input uses the corresponding
.IP
subvolume from the roi file
.IP
[\-mean] \- compute weighted mean
.IP
[\-stdev] \- compute weighted standard deviation
.IP
[\-sample] \- estimate population stdev from the sample
.IP
[\-percentile] \- compute weighted percentile
.IP
<percent> \- the percentile to find
.IP
[\-sum] \- compute weighted sum
.IP
[\-show\-map\-name] \- print map index and name before each output
.IP
For each subvolume of the input, a single number is printed, resulting
from the specified operation.  If \fB\-weight\-volume\fR is not specified, each
voxel's volume is used.  Use \fB\-subvolume\fR to only give output for a single
subvolume.  Use \fB\-roi\fR to consider only the data within a region.  Exactly
one of \fB\-mean\fR, \fB\-stdev\fR, \fB\-percentile\fR or \fB\-sum\fR must be specified.
.IP
Using \fB\-sum\fR without \fB\-weight\-volume\fR is equivalent to integrating with
respect to volume.
.PP
\fB\-wbsparse\-merge\-dense\fR
MERGE WBSPARSE FILES ALONG DENSE DIMENSION
.IP
wb_command \fB\-wbsparse\-merge\-dense\fR
.IP
<direction> \- which dimension to merge along, ROW or COLUMN
<wbsparse\-out> \- output \- the output wbsparse file
.IP
[\-wbsparse] \- repeatable \- specify an input wbsparse file
.IP
<wbsparse\-in> \- a wbsparse file to merge
.IP
The input wbsparse files must have matching mappings along the direction
not specified, and the mapping along the specified direction must be
brain models.
.PP
\fB\-zip\-scene\-file\fR
ZIP A SCENE FILE AND ITS DATA FILES
.IP
wb_command \fB\-zip\-scene\-file\fR
.IP
<scene\-file> \- the scene file to make the zip file from
<extract\-folder> \- the name of the folder created when the zip file is
.IP
unzipped
.IP
<zip\-file> \- out \- the zip file that will be created
.IP
[\-base\-dir] \- specify a directory that all data files are somewhere
.IP
within, this will become the root of the zipfile's directory structure
<directory> \- the directory
.TP
If zip\-file already exists, it will be overwritten.
If \fB\-base\-dir\fR is not
.IP
specified, the directory containing the scene file is used for the base
directory.  The scene file must contain only relative paths, and no data
files may be outside the base directory.
.PP
\fB\-zip\-spec\-file\fR
ZIP A SPEC FILE AND ITS DATA FILES
.IP
wb_command \fB\-zip\-spec\-file\fR
.IP
<spec\-file> \- the specification file to add to zip file
<extract\-folder> \- the name of the folder created when the zip file is
.IP
unzipped
.IP
<zip\-file> \- out \- the zip file that will be created
.IP
[\-base\-dir] \- specify a directory that all data files are somewhere
.IP
within, this will become the root of the zipfile's directory structure
<directory> \- the directory
.TP
If zip\-file already exists, it will be overwritten.
If \fB\-base\-dir\fR is not
.IP
specified, the directory containing the spec file is used for the base
directory.  The spec file must contain only relative paths, and no data
files may be outside the base directory.  Scene files inside spec files
are not checked for what files they reference, ensure that all data files
referenced by the scene files are also referenced by the spec file.