BEAGLE

NAME¶

libhmsbeagle/beagle.h - This file documents the API as well as header for the Broad-platform Evolutionary Analysis General Likelihood Evaluator.

SYNOPSIS¶

#include 'libhmsbeagle/platform.h'

Classes¶

struct BeagleInstanceDetails
Information about a specific instance. struct BeagleResource
Description of a hardware resource. struct BeagleResourceList
List of hardware resources. struct BeagleBenchmarkedResource
Description of a benchmarked hardware resource. struct BeagleBenchmarkedResourceList
Ordered list of benchmarked hardware resources, from fastest to slowest. struct BeagleOperation
A list of integer indices which specify a partial likelihoods operation. struct BeagleOperationByPartition
A list of integer indices which specify a partial likelihoods operation for a partitioned analysis.

Enumerations¶

enum BeagleReturnCodes { BEAGLE_SUCCESS = 0, BEAGLE_ERROR_GENERAL = -1, BEAGLE_ERROR_OUT_OF_MEMORY = -2, BEAGLE_ERROR_UNIDENTIFIED_EXCEPTION = -3, BEAGLE_ERROR_UNINITIALIZED_INSTANCE = -4, BEAGLE_ERROR_OUT_OF_RANGE = -5, BEAGLE_ERROR_NO_RESOURCE = -6, BEAGLE_ERROR_NO_IMPLEMENTATION = -7, BEAGLE_ERROR_FLOATING_POINT = -8 }
Error return codes. enum BeagleFlags { BEAGLE_FLAG_PRECISION_SINGLE = 1 << 0, BEAGLE_FLAG_PRECISION_DOUBLE = 1 << 1, BEAGLE_FLAG_COMPUTATION_SYNCH = 1 << 2, BEAGLE_FLAG_COMPUTATION_ASYNCH = 1 << 3, BEAGLE_FLAG_EIGEN_REAL = 1 << 4, BEAGLE_FLAG_EIGEN_COMPLEX = 1 << 5, BEAGLE_FLAG_SCALING_MANUAL = 1 << 6, BEAGLE_FLAG_SCALING_AUTO = 1 << 7, BEAGLE_FLAG_SCALING_ALWAYS = 1 << 8, BEAGLE_FLAG_SCALING_DYNAMIC = 1 << 25, BEAGLE_FLAG_SCALERS_RAW = 1 << 9, BEAGLE_FLAG_SCALERS_LOG = 1 << 10, BEAGLE_FLAG_INVEVEC_STANDARD = 1 << 20, BEAGLE_FLAG_INVEVEC_TRANSPOSED = 1 << 21, BEAGLE_FLAG_VECTOR_SSE = 1 << 11, BEAGLE_FLAG_VECTOR_AVX = 1 << 24, BEAGLE_FLAG_VECTOR_NONE = 1 << 12, BEAGLE_FLAG_THREADING_CPP = 1 << 30, BEAGLE_FLAG_THREADING_OPENMP = 1 << 13, BEAGLE_FLAG_THREADING_NONE = 1 << 14, BEAGLE_FLAG_PROCESSOR_CPU = 1 << 15, BEAGLE_FLAG_PROCESSOR_GPU = 1 << 16, BEAGLE_FLAG_PROCESSOR_FPGA = 1 << 17, BEAGLE_FLAG_PROCESSOR_CELL = 1 << 18, BEAGLE_FLAG_PROCESSOR_PHI = 1 << 19, BEAGLE_FLAG_PROCESSOR_OTHER = 1 << 26, BEAGLE_FLAG_FRAMEWORK_CUDA = 1 << 22, BEAGLE_FLAG_FRAMEWORK_OPENCL = 1 << 23, BEAGLE_FLAG_FRAMEWORK_CPU = 1 << 27, BEAGLE_FLAG_PARALLELOPS_STREAMS = 1 << 28, BEAGLE_FLAG_PARALLELOPS_GRID = 1 << 29 }
Hardware and implementation capability flags. enum BeagleBenchmarkFlags { BEAGLE_BENCHFLAG_SCALING_NONE = 1 << 0, BEAGLE_BENCHFLAG_SCALING_ALWAYS = 1 << 1, BEAGLE_BENCHFLAG_SCALING_DYNAMIC = 1 << 2 }
Benchmarking mode flags for resource performance evaluation with beagleGetOrderedResourceList. enum BeagleOpCodes { BEAGLE_OP_COUNT = 7, BEAGLE_PARTITION_OP_COUNT = 9, BEAGLE_OP_NONE = -1 }
Operation codes.

Functions¶

BEAGLE_DLLEXPORT const char * beagleGetVersion (void)
Get version. BEAGLE_DLLEXPORT const char * beagleGetCitation (void)
Get citation. BEAGLE_DLLEXPORT BeagleResourceList * beagleGetResourceList (void)
Get list of hardware resources. BEAGLE_DLLEXPORT BeagleBenchmarkedResourceList * beagleGetBenchmarkedResourceList (int tipCount, int compactBufferCount, int stateCount, int patternCount, int categoryCount, int *resourceList, int resourceCount, long preferenceFlags, long requirementFlags, int eigenModelCount, int partitionCount, int calculateDerivatives, long benchmarkFlags)
Get a benchmarked list of hardware resources for the given analysis parameters. BEAGLE_DLLEXPORT int beagleCreateInstance (int tipCount, int partialsBufferCount, int compactBufferCount, int stateCount, int patternCount, int eigenBufferCount, int matrixBufferCount, int categoryCount, int scaleBufferCount, int *resourceList, int resourceCount, long preferenceFlags, long requirementFlags, BeagleInstanceDetails *returnInfo)
Create a single instance. BEAGLE_DLLEXPORT int beagleFinalizeInstance (int instance)
Finalize this instance. BEAGLE_DLLEXPORT int beagleFinalize (void)
Finalize the library. BEAGLE_DLLEXPORT int beagleSetCPUThreadCount (int instance, int threadCount)
Set number of threads for native CPU implementation. BEAGLE_DLLEXPORT int beagleSetTipStates (int instance, int tipIndex, const int *inStates)
Set the compact state representation for tip node. BEAGLE_DLLEXPORT int beagleSetTipPartials (int instance, int tipIndex, const double *inPartials)
Set an instance partials buffer for tip node. BEAGLE_DLLEXPORT int beagleSetPartials (int instance, int bufferIndex, const double *inPartials)
Set an instance partials buffer. BEAGLE_DLLEXPORT int beagleGetPartials (int instance, int bufferIndex, int scaleIndex, double *outPartials)
Get partials from an instance buffer. BEAGLE_DLLEXPORT int beagleSetEigenDecomposition (int instance, int eigenIndex, const double *inEigenVectors, const double *inInverseEigenVectors, const double *inEigenValues)
Set an eigen-decomposition buffer. BEAGLE_DLLEXPORT int beagleSetStateFrequencies (int instance, int stateFrequenciesIndex, const double *inStateFrequencies)
Set a state frequency buffer. BEAGLE_DLLEXPORT int beagleSetCategoryWeights (int instance, int categoryWeightsIndex, const double *inCategoryWeights)
Set a category weights buffer. BEAGLE_DLLEXPORT int beagleSetCategoryRates (int instance, const double *inCategoryRates)
Set the default category rates buffer. BEAGLE_DLLEXPORT int beagleSetCategoryRatesWithIndex (int instance, int categoryRatesIndex, const double *inCategoryRates)
Set a category rates buffer. BEAGLE_DLLEXPORT int beagleSetPatternWeights (int instance, const double *inPatternWeights)
Set pattern weights. BEAGLE_DLLEXPORT int beagleSetPatternPartitions (int instance, int partitionCount, const int *inPatternPartitions)
Set pattern partition assignments. BEAGLE_DLLEXPORT int beagleConvolveTransitionMatrices (int instance, const int *firstIndices, const int *secondIndices, const int *resultIndices, int matrixCount)
Set partitions by pattern weight. BEAGLE_DLLEXPORT int beagleUpdateTransitionMatrices (int instance, int eigenIndex, const int *probabilityIndices, const int *firstDerivativeIndices, const int *secondDerivativeIndices, const double *edgeLengths, int count)
Calculate a list of transition probability matrices. BEAGLE_DLLEXPORT int beagleUpdateTransitionMatricesWithMultipleModels (int instance, const int *eigenIndices, const int *categoryRateIndices, const int *probabilityIndices, const int *firstDerivativeIndices, const int *secondDerivativeIndices, const double *edgeLengths, int count)
Calculate a list of transition probability matrices with multiple models. BEAGLE_DLLEXPORT int beagleSetTransitionMatrix (int instance, int matrixIndex, const double *inMatrix, double paddedValue)
Set a finite-time transition probability matrix. BEAGLE_DLLEXPORT int beagleGetTransitionMatrix (int instance, int matrixIndex, double *outMatrix)
Get a finite-time transition probability matrix. BEAGLE_DLLEXPORT int beagleSetTransitionMatrices (int instance, const int *matrixIndices, const double *inMatrices, const double *paddedValues, int count)
Set multiple transition matrices. BEAGLE_DLLEXPORT int beagleUpdatePartials (const int instance, const BeagleOperation *operations, int operationCount, int cumulativeScaleIndex)
Calculate or queue for calculation partials using a list of operations. BEAGLE_DLLEXPORT int beagleUpdatePartialsByPartition (const int instance, const BeagleOperationByPartition *operations, int operationCount)
Calculate or queue for calculation partials using a list of partition operations. BEAGLE_DLLEXPORT int beagleWaitForPartials (const int instance, const int *destinationPartials, int destinationPartialsCount)
Block until all calculations that write to the specified partials have completed. BEAGLE_DLLEXPORT int beagleAccumulateScaleFactors (int instance, const int *scaleIndices, int count, int cumulativeScaleIndex)
Accumulate scale factors. BEAGLE_DLLEXPORT int beagleAccumulateScaleFactorsByPartition (int instance, const int *scaleIndices, int count, int cumulativeScaleIndex, int partitionIndex)
Accumulate scale factors by partition. BEAGLE_DLLEXPORT int beagleRemoveScaleFactors (int instance, const int *scaleIndices, int count, int cumulativeScaleIndex)
Remove scale factors. BEAGLE_DLLEXPORT int beagleRemoveScaleFactorsByPartition (int instance, const int *scaleIndices, int count, int cumulativeScaleIndex, int partitionIndex)
Remove scale factors by partition. BEAGLE_DLLEXPORT int beagleResetScaleFactors (int instance, int cumulativeScaleIndex)
Reset scalefactors. BEAGLE_DLLEXPORT int beagleResetScaleFactorsByPartition (int instance, int cumulativeScaleIndex, int partitionIndex)
Reset scalefactors by partition. BEAGLE_DLLEXPORT int beagleCopyScaleFactors (int instance, int destScalingIndex, int srcScalingIndex)
Copy scale factors. BEAGLE_DLLEXPORT int beagleGetScaleFactors (int instance, int srcScalingIndex, double *outScaleFactors)
Get scale factors. BEAGLE_DLLEXPORT int beagleCalculateRootLogLikelihoods (int instance, const int *bufferIndices, const int *categoryWeightsIndices, const int *stateFrequenciesIndices, const int *cumulativeScaleIndices, int count, double *outSumLogLikelihood)
Calculate site log likelihoods at a root node. BEAGLE_DLLEXPORT int beagleCalculateRootLogLikelihoodsByPartition (int instance, const int *bufferIndices, const int *categoryWeightsIndices, const int *stateFrequenciesIndices, const int *cumulativeScaleIndices, const int *partitionIndices, int partitionCount, int count, double *outSumLogLikelihoodByPartition, double *outSumLogLikelihood)
Calculate site log likelihoods at a root node with per partition buffers. BEAGLE_DLLEXPORT int beagleCalculateEdgeLogLikelihoods (int instance, const int *parentBufferIndices, const int *childBufferIndices, const int *probabilityIndices, const int *firstDerivativeIndices, const int *secondDerivativeIndices, const int *categoryWeightsIndices, const int *stateFrequenciesIndices, const int *cumulativeScaleIndices, int count, double *outSumLogLikelihood, double *outSumFirstDerivative, double *outSumSecondDerivative)
Calculate site log likelihoods and derivatives along an edge. BEAGLE_DLLEXPORT int beagleCalculateEdgeLogLikelihoodsByPartition (int instance, const int *parentBufferIndices, const int *childBufferIndices, const int *probabilityIndices, const int *firstDerivativeIndices, const int *secondDerivativeIndices, const int *categoryWeightsIndices, const int *stateFrequenciesIndices, const int *cumulativeScaleIndices, const int *partitionIndices, int partitionCount, int count, double *outSumLogLikelihoodByPartition, double *outSumLogLikelihood, double *outSumFirstDerivativeByPartition, double *outSumFirstDerivative, double *outSumSecondDerivativeByPartition, double *outSumSecondDerivative)
Calculate multiple site log likelihoods and derivatives along an edge with per partition buffers. BEAGLE_DLLEXPORT int beagleGetLogLikelihood (int instance, double *outSumLogLikelihood)
Returns log likelihood sum and subsequent to an asynchronous integration call. BEAGLE_DLLEXPORT int beagleGetDerivatives (int instance, double *outSumFirstDerivative, double *outSumSecondDerivative)
Returns derivative sums subsequent to an asynchronous integration call. BEAGLE_DLLEXPORT int beagleGetSiteLogLikelihoods (int instance, double *outLogLikelihoods)
Get site log likelihoods for last beagleCalculateRootLogLikelihoods or beagleCalculateEdgeLogLikelihoods call. BEAGLE_DLLEXPORT int beagleGetSiteDerivatives (int instance, double *outFirstDerivatives, double *outSecondDerivatives)
Get site derivatives for last beagleCalculateEdgeLogLikelihoods call.

Detailed Description¶

This file documents the API as well as header for the Broad-platform Evolutionary Analysis General Likelihood Evaluator.

This file is part of BEAGLE.

BEAGLE is free software: you can redistribute it and/or modify it under the terms of the GNU Lesser General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.

BEAGLE is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License for more details.

You should have received a copy of the GNU Lesser General Public License along with BEAGLE. If not, see http://www.gnu.org/licenses/.

KEY CONCEPTS

The key to BEAGLE performance lies in delivering fine-scale parallelization while minimizing data transfer and memory copy overhead. To accomplish this, the library lacks the concept of data structure for a tree, in spite of the intended use for phylogenetic analysis. Instead, BEAGLE acts directly on flexibly indexed data storage (called buffers) for observed character states and partial likelihoods. The client program can set the input buffers to reflect the data and can calculate the likelihood of a particular phylogeny by invoking likelihood calculations on the appropriate input and output buffers in the correct order. Because of this design simplicity, the library can support many different tree inference algorithms and likelihood calculation on a variety of models. Arbitrary numbers of states can be used, as can nonreversible substitution matrices via complex eigen decompositions, and mixture models with multiple rate categories and/or multiple eigen decompositions. Finally, BEAGLE application programming interface (API) calls can be asynchronous, allowing the calling program to implement other coarse-scale parallelization schemes such as evaluating independent genes or running concurrent Markov chains.

USAGE

To use the library, a client program first creates an instance of BEAGLE by calling beagleCreateInstance; multiple instances per client are possible and encouraged. All additional functions are called with a reference to this instance. The client program can optionally request that an instance run on certain hardware (e.g., a GPU) or have particular features (e.g., double-precision math). Next, the client program must specify the data dimensions and specify key aspects of the phylogenetic model. Character state data are then loaded and can be in the form of discrete observed states or partial likelihoods for ambiguous characters. The observed data are usually unchanging and loaded only once at the start to minimize memory copy overhead. The character data can be compressed into unique “site patterns” and associated weights for each. The parameters of the substitution process can then be specified, including the equilibrium state frequencies, the rates for one or more substitution rate categories and their weights, and finally, the eigen decomposition for the substitution process.

In order to calculate the likelihood of a particular tree, the client program then specifies a series of integration operations that correspond to steps in Felsenstein’s algorithm. Finite-time transition probabilities for each edge are loaded directly if considering a nondiagonalizable model or calculated in parallel from the eigen decomposition and edge lengths specified. This is performed within BEAGLE’s memory space to minimize data transfers. A single function call will then request one or more integration operations to calculate partial likelihoods over some or all nodes. The operations are performed in the order they are provided, typically dictated by a postorder traversal of the tree topology. The client needs only specify nodes for which the partial likelihoods need updating, but it is up to the calling software to keep track of these dependencies. The final step in evaluating the phylogenetic model is done using an API call that yields a single log likelihood for the model given the data.

Aspects of the BEAGLE API design support both maximum likelihood (ML) and Bayesian phylogenetic tree inference. For ML inference, API calls can calculate first and second derivatives of the likelihood with respect to the lengths of edges (branches). In both cases, BEAGLE provides the ability to cache and reuse previously computed partial likelihood results, which can yield a tremendous speedup over recomputing the entire likelihood every time a new phylogenetic model is evaluated.

Author

Likelihood API Working Group

Daniel Ayres

Peter Beerli

Michael Cummings

Aaron Darling

Mark Holder

John Huelsenbeck

Paul Lewis

Michael Ott

Andrew Rambaut

Fredrik Ronquist

Marc Suchard

David Swofford

Derrick Zwickl

Enumeration Type Documentation¶

enum BeagleBenchmarkFlags¶

Benchmarking mode flags for resource performance evaluation with beagleGetOrderedResourceList.

This enumerates all possible benchmarking mode flags. Each mode is a bit in a 'long'

Enumerator

BEAGLE_BENCHFLAG_SCALING_NONE: No scaling
BEAGLE_BENCHFLAG_SCALING_ALWAYS: Scale at every iteration
BEAGLE_BENCHFLAG_SCALING_DYNAMIC: Scale every fixed number of iterations or when a numerical error occurs, and re-use scale factors for subsequent iterations

enum BeagleFlags¶

Hardware and implementation capability flags.

This enumerates all possible hardware and implementation capability flags. Each capability is a bit in a 'long'

Enumerator

BEAGLE_FLAG_PRECISION_SINGLE: Single precision computation
BEAGLE_FLAG_PRECISION_DOUBLE: Double precision computation
BEAGLE_FLAG_COMPUTATION_SYNCH: Synchronous computation (blocking)
BEAGLE_FLAG_COMPUTATION_ASYNCH: Asynchronous computation (non-blocking)
BEAGLE_FLAG_EIGEN_REAL: Real eigenvalue computation
BEAGLE_FLAG_EIGEN_COMPLEX: Complex eigenvalue computation
BEAGLE_FLAG_SCALING_MANUAL: Manual scaling
BEAGLE_FLAG_SCALING_AUTO: Auto-scaling on (deprecated, may not work correctly)
BEAGLE_FLAG_SCALING_ALWAYS: Scale at every updatePartials (deprecated, may not work correctly)
BEAGLE_FLAG_SCALING_DYNAMIC: Manual scaling with dynamic checking (deprecated, may not work correctly)
BEAGLE_FLAG_SCALERS_RAW: Save raw scalers
BEAGLE_FLAG_SCALERS_LOG: Save log scalers
BEAGLE_FLAG_INVEVEC_STANDARD: Inverse eigen vectors passed to BEAGLE have not been transposed
BEAGLE_FLAG_INVEVEC_TRANSPOSED: Inverse eigen vectors passed to BEAGLE have been transposed
BEAGLE_FLAG_VECTOR_SSE: SSE computation
BEAGLE_FLAG_VECTOR_AVX: AVX computation
BEAGLE_FLAG_VECTOR_NONE: No vector computation
BEAGLE_FLAG_THREADING_CPP: C++11 threading
BEAGLE_FLAG_THREADING_OPENMP: OpenMP threading
BEAGLE_FLAG_THREADING_NONE: No threading (default)
BEAGLE_FLAG_PROCESSOR_CPU: Use CPU as main processor
BEAGLE_FLAG_PROCESSOR_GPU: Use GPU as main processor
BEAGLE_FLAG_PROCESSOR_FPGA: Use FPGA as main processor
BEAGLE_FLAG_PROCESSOR_CELL: Use Cell as main processor
BEAGLE_FLAG_PROCESSOR_PHI: Use Intel Phi as main processor
BEAGLE_FLAG_PROCESSOR_OTHER: Use other type of processor
BEAGLE_FLAG_FRAMEWORK_CUDA: Use CUDA implementation with GPU resources
BEAGLE_FLAG_FRAMEWORK_OPENCL: Use OpenCL implementation with GPU resources
BEAGLE_FLAG_FRAMEWORK_CPU: Use CPU implementation
BEAGLE_FLAG_PARALLELOPS_STREAMS: Operations in updatePartials may be assigned to separate device streams
BEAGLE_FLAG_PARALLELOPS_GRID: Operations in updatePartials may be folded into single kernel launch (necessary for partitions; typically performs better for problems with fewer pattern sites)

enum BeagleOpCodes¶

Operation codes.

This enumerates all possible BEAGLE operation codes.

Enumerator

BEAGLE_OP_COUNT: Total number of integers per beagleUpdatePartials operation
BEAGLE_PARTITION_OP_COUNT: Total number of integers per beagleUpdatePartialsByPartition operation
BEAGLE_OP_NONE: Specify no use for indexed buffer

enum BeagleReturnCodes¶

Error return codes.

This enumerates all possible BEAGLE return codes. Error codes are always negative.

Enumerator

BEAGLE_SUCCESS: Success
BEAGLE_ERROR_GENERAL: Unspecified error
BEAGLE_ERROR_OUT_OF_MEMORY: Not enough memory could be allocated
BEAGLE_ERROR_UNIDENTIFIED_EXCEPTION: Unspecified exception
BEAGLE_ERROR_UNINITIALIZED_INSTANCE: The instance index is out of range, or the instance has not been created
BEAGLE_ERROR_OUT_OF_RANGE: One of the indices specified exceeded the range of the array
BEAGLE_ERROR_NO_RESOURCE: No resource matches requirements
BEAGLE_ERROR_NO_IMPLEMENTATION: No implementation matches requirements
BEAGLE_ERROR_FLOATING_POINT: Floating-point error (e.g., NaN)

Function Documentation¶

BEAGLE_DLLEXPORT int beagleAccumulateScaleFactors (int instance, const int * scaleIndices, int count, int cumulativeScaleIndex)¶

Accumulate scale factors. This function adds (log) scale factors from a list of scaleBuffers to a cumulative scale buffer. It is used to calculate the marginal scaling at a specific node for each site.

Parameters

instance Instance number (input)
scaleIndices List of scaleBuffers to add (input)
count Number of scaleBuffers in list (input)
cumulativeScaleIndex Index number of scaleBuffer to accumulate factors into (input)

BEAGLE_DLLEXPORT int beagleAccumulateScaleFactorsByPartition (int instance, const int * scaleIndices, int count, int cumulativeScaleIndex, int partitionIndex)¶

Accumulate scale factors by partition. This function adds (log) scale factors from a list of scaleBuffers to a cumulative scale buffer. It is used to calculate the marginal scaling at a specific node for each site.

Parameters

BEAGLE_DLLEXPORT int beagleCalculateEdgeLogLikelihoods (int instance, const int * parentBufferIndices, const int * childBufferIndices, const int * probabilityIndices, const int * firstDerivativeIndices, const int * secondDerivativeIndices, const int * categoryWeightsIndices, const int * stateFrequenciesIndices, const int * cumulativeScaleIndices, int count, double * outSumLogLikelihood, double * outSumFirstDerivative, double * outSumSecondDerivative)¶

Calculate site log likelihoods and derivatives along an edge. This function integrates a list of partials at a parent and child node with respect to a set of partials-weights and state frequencies to return the log likelihood and first and second derivative sums

Parameters

Returns

error code

BEAGLE_DLLEXPORT int beagleCalculateEdgeLogLikelihoodsByPartition (int instance, const int * parentBufferIndices, const int * childBufferIndices, const int * probabilityIndices, const int * firstDerivativeIndices, const int * secondDerivativeIndices, const int * categoryWeightsIndices, const int * stateFrequenciesIndices, const int * cumulativeScaleIndices, const int * partitionIndices, int partitionCount, int count, double * outSumLogLikelihoodByPartition, double * outSumLogLikelihood, double * outSumFirstDerivativeByPartition, double * outSumFirstDerivative, double * outSumSecondDerivativeByPartition, double * outSumSecondDerivative)¶

Calculate multiple site log likelihoods and derivatives along an edge with per partition buffers. This function integrates lists of partials at a parent and child node with respect to a set of partials-weights and state frequencies to return the log likelihood and first and second derivative sums

Parameters

instance Instance number (input)
parentBufferIndices List of indices of parent partialsBuffers (input)
childBufferIndices List of indices of child partialsBuffers (input)
probabilityIndices List indices of transition probability matrices for this edge (input)
firstDerivativeIndices List indices of first derivative matrices (input)
secondDerivativeIndices List indices of second derivative matrices (input)
categoryWeightsIndices List of weights to apply to each partialsBuffer (input)
stateFrequenciesIndices List of state frequencies for each partialsBuffer (input). There should be one set for each of parentBufferIndices
cumulativeScaleIndices List of scaleBuffers containing accumulated factors to apply to each partialsBuffer (input). There should be one index for each of parentBufferIndices
partitionIndices List of partition indices indicating which sites in each partialsBuffer should be used (input). There should be one index for each of parentBufferIndices
partitionCount Number of distinct partitionIndices (input)
count Number of sets of partitions to integrate across (input)
outSumLogLikelihoodByPartition Pointer to destination for resulting log likelihoods for each partition (output)
outSumLogLikelihood Pointer to destination for resulting log likelihood (output)
outSumFirstDerivativeByPartition Pointer to destination for resulting first derivative for each partition (output)
outSumFirstDerivative Pointer to destination for resulting first derivative (output)
outSumSecondDerivativeByPartition Pointer to destination for resulting second derivative for each partition (output)
outSumSecondDerivative Pointer to destination for resulting second derivative (output)

Returns

error code

BEAGLE_DLLEXPORT int beagleCalculateRootLogLikelihoods (int instance, const int * bufferIndices, const int * categoryWeightsIndices, const int * stateFrequenciesIndices, const int * cumulativeScaleIndices, int count, double * outSumLogLikelihood)¶

Calculate site log likelihoods at a root node. This function integrates a list of partials at a node with respect to a set of partials-weights and state frequencies to return the log likelihood sum

Parameters

Returns

error code

BEAGLE_DLLEXPORT int beagleCalculateRootLogLikelihoodsByPartition (int instance, const int * bufferIndices, const int * categoryWeightsIndices, const int * stateFrequenciesIndices, const int * cumulativeScaleIndices, const int * partitionIndices, int partitionCount, int count, double * outSumLogLikelihoodByPartition, double * outSumLogLikelihood)¶

Calculate site log likelihoods at a root node with per partition buffers. This function integrates lists of partials at a node with respect to a set of partials-weights and state frequencies to return the log likelihood sums

Parameters

instance Instance number (input)
bufferIndices List of partialsBuffer indices to integrate (input)
categoryWeightsIndices List of weights to apply to each partialsBuffer (input). There should be one categoryCount sized set for each of bufferIndices
stateFrequenciesIndices List of state frequencies for each partialsBuffer (input). There should be one set for each of bufferIndices
cumulativeScaleIndices List of scaleBuffers containing accumulated factors to apply to each partialsBuffer (input). There should be one index for each of bufferIndices
partitionIndices List of partition indices indicating which sites in each partialsBuffer should be used (input). There should be one index for each of bufferIndices
partitionCount Number of distinct partitionIndices (input)
count Number of sets of partitions to integrate across (input)
outSumLogLikelihoodByPartition Pointer to destination for resulting log likelihoods for each partition (output)
outSumLogLikelihood Pointer to destination for resulting log likelihood (output)

Returns

error code

BEAGLE_DLLEXPORT int beagleConvolveTransitionMatrices (int instance, const int * firstIndices, const int * secondIndices, const int * resultIndices, int matrixCount)¶

Set partitions by pattern weight. This function defines partitions by setting their vectors of pattern-weights for an instance. It should be called after beagleSetTipPartials or beagleSetTipStates.

Parameters

instance Instance number (input)
inPartitionPatternWeights partitionCount arrays, each containing patternCount pattern weights (input)
partitionCount Number of partitions (input)

Returns

error code

Convolve lists of transition probability matrices

This function convolves two lists of transition probability matrices.

Parameters

instance Instance number (input)
firstIndices List of indices of the first transition probability matrices to convolve (input)
secondIndices List of indices of the second transition probability matrices to convolve (input)
resultIndices List of indices of resulting transition probability matrices (input)
matrixCount Length of lists

BEAGLE_DLLEXPORT int beagleCopyScaleFactors (int instance, int destScalingIndex, int srcScalingIndex)¶

Copy scale factors. This function copies scale factors from one buffer to another.

Parameters

instance Instance number (input)
destScalingIndex Destination scaleBuffer (input)
srcScalingIndex Source scaleBuffer (input)

BEAGLE_DLLEXPORT int beagleCreateInstance (int tipCount, int partialsBufferCount, int compactBufferCount, int stateCount, int patternCount, int eigenBufferCount, int matrixBufferCount, int categoryCount, int scaleBufferCount, int * resourceList, int resourceCount, long preferenceFlags, long requirementFlags, BeagleInstanceDetails * returnInfo)¶

Create a single instance. This function creates a single instance of the BEAGLE library and can be called multiple times to create multiple data partition instances each returning a unique identifier.

Parameters

tipCount Number of tip data elements (input)
partialsBufferCount Number of partials buffers to create (input)
compactBufferCount Number of compact state representation buffers to create (input)
stateCount Number of states in the continuous-time Markov chain (input)
patternCount Number of site patterns to be handled by the instance (input)
eigenBufferCount Number of rate matrix eigen-decomposition, category weight, category rates, and state frequency buffers to allocate (input)
matrixBufferCount Number of transition probability matrix buffers (input)
categoryCount Number of rate categories (input)
scaleBufferCount Number of scale buffers to create, ignored for auto scale or always scale (input)
resourceList List of potential resources on which this instance is allowed (input, NULL implies no restriction)
resourceCount Length of resourceList list (input)
preferenceFlags Bit-flags indicating preferred implementation characteristics, see BeagleFlags (input)
requirementFlags Bit-flags indicating required implementation characteristics, see BeagleFlags (input)
returnInfo Pointer to return implementation and resource details

Returns

the unique instance identifier (<0 if failed, see BeagleReturnCodes)

BEAGLE_DLLEXPORT int beagleFinalize (void)¶

Finalize the library. This function finalizes the library and releases all allocated memory. This function is automatically called under GNU C via attribute ((destructor)).

Returns

error code

BEAGLE_DLLEXPORT int beagleFinalizeInstance (int instance)¶

Finalize this instance. This function finalizes the instance by releasing allocated memory

Parameters

instance Instance number

Returns

error code

BEAGLE_DLLEXPORT BeagleBenchmarkedResourceList* beagleGetBenchmarkedResourceList (int tipCount, int compactBufferCount, int stateCount, int patternCount, int categoryCount, int * resourceList, int resourceCount, long preferenceFlags, long requirementFlags, int eigenModelCount, int partitionCount, int calculateDerivatives, long benchmarkFlags)¶

Get a benchmarked list of hardware resources for the given analysis parameters. This function returns a pointer to a BeagleBenchmarkedResourceList struct, which includes a BeagleBenchmarkedResource array describing the available hardware resources with benchmark times and CPU performance ratios for each resource. Resources are benchmarked with the given analysis parameters and the array is ordered from fastest to slowest. If there is an error the function returns NULL.

Parameters

tipCount Number of tip data elements (input)
compactBufferCount Number of compact state representation tips (input)
stateCount Number of states in the continuous-time Markov chain (input)
patternCount Number of site patterns (input)
categoryCount Number of rate categories (input)
resourceList List of resources to be benchmarked, NULL implies no restriction (input)
resourceCount Length of resourceList list (input)
preferenceFlags Bit-flags indicating preferred implementation characteristics, see BeagleFlags (input)
requirementFlags Bit-flags indicating required implementation characteristics, see BeagleFlags (input)
eigenModelCount Number of full-alignment rate matrix eigen-decomposition models (input)
partitionCount Number of partitions (input)
calculateDerivatives Indicates if calculation of derivatives are required (input)
benchmarkFlags Bit-flags indicating benchmarking preferences (input)

Returns

An ordered (fastest to slowest) list of hardware resources available to the library as a BeagleBenchmarkedResourceList for the specified analysis parameters

BEAGLE_DLLEXPORT const char* beagleGetCitation (void)¶

Get citation. This function returns a pointer to a string describing the version of the library and how to cite it.

Returns

A string describing the version of the library and how to cite it

BEAGLE_DLLEXPORT int beagleGetDerivatives (int instance, double * outSumFirstDerivative, double * outSumSecondDerivative)¶

Returns derivative sums subsequent to an asynchronous integration call. This function is optional and only has to be called by clients that use the non-blocking asynchronous computation mode (BEAGLE_FLAG_COMPUTATION_ASYNCH).

If used, this function must be called after a beagleCalculateEdgeLogLikelihoods call. The library will block until the derivatiives have been calculated.

Parameters

instance Instance number (input)
outSumFirstDerivative Pointer to destination for resulting first derivative (output)
outSumSecondDerivative Pointer to destination for resulting second derivative (output)

Returns

error code

BEAGLE_DLLEXPORT int beagleGetLogLikelihood (int instance, double * outSumLogLikelihood)¶

Returns log likelihood sum and subsequent to an asynchronous integration call. This function is optional and only has to be called by clients that use the non-blocking asynchronous computation mode (BEAGLE_FLAG_COMPUTATION_ASYNCH).

If used, this function must be called after a beagleCalculateRootLogLikelihoods or beagleCalculateEdgeLogLikelihoods call. The library will block until the likelihood has been calculated.

Parameters

instance Instance number (input)
outSumLogLikelihood Pointer to destination for resulting log likelihood (output)

Returns

error code

BEAGLE_DLLEXPORT int beagleGetPartials (int instance, int bufferIndex, int scaleIndex, double * outPartials)¶

Get partials from an instance buffer. This function copies an instance buffer into the array outPartials. The outPartials array should be stateCount * patternCount * categoryCount in length.

Parameters

instance Instance number from which to get partialsBuffer (input)
bufferIndex Index of source partialsBuffer (input)
scaleIndex Index of scaleBuffer to apply to partialsBuffer (input)
outPartials Pointer to which to receive partialsBuffer (output)

Returns

error code

BEAGLE_DLLEXPORT BeagleResourceList* beagleGetResourceList (void)¶

Get list of hardware resources. This function returns a pointer to a BeagleResourceList struct, which includes a BeagleResource array describing the available hardware resources.

Returns

A list of hardware resources available to the library as a BeagleResourceList

BEAGLE_DLLEXPORT int beagleGetScaleFactors (int instance, int srcScalingIndex, double * outScaleFactors)¶

Get scale factors. This function retrieves a buffer of scale factors.

Parameters

instance Instance number (input)
srcScalingIndex Source scaleBuffer (input)
outScaleFactors Pointer to which to receive scaleFactors (output)

BEAGLE_DLLEXPORT int beagleGetSiteDerivatives (int instance, double * outFirstDerivatives, double * outSecondDerivatives)¶

Get site derivatives for last beagleCalculateEdgeLogLikelihoods call. This function returns the derivatives for each site

Parameters

instance Instance number (input)
outFirstDerivatives Pointer to destination for resulting first derivatives (output)
outSecondDerivatives Pointer to destination for resulting second derivatives (output)

Returns

error code

BEAGLE_DLLEXPORT int beagleGetSiteLogLikelihoods (int instance, double * outLogLikelihoods)¶

Get site log likelihoods for last beagleCalculateRootLogLikelihoods or beagleCalculateEdgeLogLikelihoods call. This function returns the log likelihoods for each site

Parameters

instance Instance number (input)
outLogLikelihoods Pointer to destination for resulting log likelihoods (output)

Returns

error code

BEAGLE_DLLEXPORT int beagleGetTransitionMatrix (int instance, int matrixIndex, double * outMatrix)¶

Get a finite-time transition probability matrix. This function copies a finite-time transition matrix buffer into the array outMatrix. The outMatrix array should be of size stateCount * stateCount * categoryCount and will be filled with one matrix for each rate category.

Parameters

instance Instance number (input)
matrixIndex Index of matrix buffer (input)
outMatrix Pointer to destination transition probability matrix (output)

Returns

error code

BEAGLE_DLLEXPORT const char* beagleGetVersion (void)¶

Get version. This function returns a pointer to a string with the library version number.

Returns

A string with the version number

BEAGLE_DLLEXPORT int beagleRemoveScaleFactors (int instance, const int * scaleIndices, int count, int cumulativeScaleIndex)¶

Remove scale factors. This function removes (log) scale factors from a cumulative scale buffer. The scale factors to be removed are indicated in a list of scaleBuffers.

Parameters

instance Instance number (input)
scaleIndices List of scaleBuffers to remove (input)
count Number of scaleBuffers in list (input)
cumulativeScaleIndex Index number of scaleBuffer containing accumulated factors (input)

BEAGLE_DLLEXPORT int beagleRemoveScaleFactorsByPartition (int instance, const int * scaleIndices, int count, int cumulativeScaleIndex, int partitionIndex)¶

Remove scale factors by partition. This function removes (log) scale factors from a cumulative scale buffer. The scale factors to be removed are indicated in a list of scaleBuffers.

Parameters

BEAGLE_DLLEXPORT int beagleResetScaleFactors (int instance, int cumulativeScaleIndex)¶

Reset scalefactors. This function resets a cumulative scale buffer.

Parameters

instance Instance number (input)
cumulativeScaleIndex Index number of cumulative scaleBuffer (input)

BEAGLE_DLLEXPORT int beagleResetScaleFactorsByPartition (int instance, int cumulativeScaleIndex, int partitionIndex)¶

Reset scalefactors by partition. This function resets a cumulative scale buffer.

Parameters

instance Instance number (input)
cumulativeScaleIndex Index number of cumulative scaleBuffer (input)
partitionIndex Index of partition to reset (input)

BEAGLE_DLLEXPORT int beagleSetCategoryRates (int instance, const double * inCategoryRates)¶

Set the default category rates buffer. This function sets the default vector of category rates for an instance.

Parameters

instance Instance number (input)
inCategoryRates Array containing categoryCount rate scalers (input)

Returns

error code

BEAGLE_DLLEXPORT int beagleSetCategoryRatesWithIndex (int instance, int categoryRatesIndex, const double * inCategoryRates)¶

Set a category rates buffer. This function sets the vector of category rates for a given buffer in an instance.

Parameters

instance Instance number (input)
categoryRatesIndex Index of category rates buffer (input)
inCategoryRates Array containing categoryCount rate scalers (input)

Returns

error code

BEAGLE_DLLEXPORT int beagleSetCategoryWeights (int instance, int categoryWeightsIndex, const double * inCategoryWeights)¶

Set a category weights buffer. This function copies a category weights array into an instance buffer.

Parameters

instance Instance number (input)
categoryWeightsIndex Index of category weights buffer (input)
inCategoryWeights Category weights array (categoryCount) (input)

Returns

error code

BEAGLE_DLLEXPORT int beagleSetCPUThreadCount (int instance, int threadCount)¶

Set number of threads for native CPU implementation. This function sets the max number of worker threads to be used with a native CPU implementation. It should only be called after beagleCreateInstance and requires the BEAGLE_FLAG_THREADING_CPP flag to be set. It has no effect on GPU-based implementations. It has no effect with the default BEAGLE_FLAG_THREADING_NONE setting. If BEAGLE_FLAG_THREADING_CPP is set and this function is not called BEAGLE will use a heuristic to set an appropriate number of threads.

Parameters

instance Instance number (input)
threadCount Number of threads (input)

Returns

error code

BEAGLE_DLLEXPORT int beagleSetEigenDecomposition (int instance, int eigenIndex, const double * inEigenVectors, const double * inInverseEigenVectors, const double * inEigenValues)¶

Set an eigen-decomposition buffer. This function copies an eigen-decomposition into an instance buffer.

Parameters

instance Instance number (input)
eigenIndex Index of eigen-decomposition buffer (input)
inEigenVectors Flattened matrix (stateCount x stateCount) of eigen-vectors (input)
inInverseEigenVectors Flattened matrix (stateCount x stateCount) of inverse-eigen- vectors (input)
inEigenValues Vector of eigenvalues

Returns

error code

BEAGLE_DLLEXPORT int beagleSetPartials (int instance, int bufferIndex, const double * inPartials)¶

Set an instance partials buffer. This function copies an array of partials into an instance buffer. The inPartials array should be stateCount * patternCount * categoryCount in length.

Parameters

instance Instance number in which to set a partialsBuffer (input)
bufferIndex Index of destination partialsBuffer (input)
inPartials Pointer to partials values to set (input)

Returns

error code

BEAGLE_DLLEXPORT int beagleSetPatternPartitions (int instance, int partitionCount, const int * inPatternPartitions)¶

Set pattern partition assignments. This function sets the vector of pattern partition indices for an instance. It should only be called after beagleSetTipPartials and beagleSetPatternWeights.

Parameters

instance Instance number (input)
partitionCount Number of partitions (input)
inPatternPartitions Array containing partitionCount partition indices (input)

Returns

error code

BEAGLE_DLLEXPORT int beagleSetPatternWeights (int instance, const double * inPatternWeights)¶

Set pattern weights. This function sets the vector of pattern weights for an instance.

Parameters

instance Instance number (input)
inPatternWeights Array containing patternCount weights (input)

Returns

error code

BEAGLE_DLLEXPORT int beagleSetStateFrequencies (int instance, int stateFrequenciesIndex, const double * inStateFrequencies)¶

Set a state frequency buffer. This function copies a state frequency array into an instance buffer.

Parameters

instance Instance number (input)
stateFrequenciesIndex Index of state frequencies buffer (input)
inStateFrequencies State frequencies array (stateCount) (input)

Returns

error code

BEAGLE_DLLEXPORT int beagleSetTipPartials (int instance, int tipIndex, const double * inPartials)¶

Set an instance partials buffer for tip node. This function copies an array of partials into an instance buffer. The inPartials array should be stateCount * patternCount in length. For most applications this will be used to set the partial likelihoods for the observed states. Internally, the partials will be copied categoryCount times.

Parameters

instance Instance number in which to set a partialsBuffer (input)
tipIndex Index of destination partialsBuffer (input)
inPartials Pointer to partials values to set (input)

Returns

error code

BEAGLE_DLLEXPORT int beagleSetTipStates (int instance, int tipIndex, const int * inStates)¶

Set the compact state representation for tip node. This function copies a compact state representation into an instance buffer. Compact state representation is an array of states: 0 to stateCount - 1 (missing = stateCount). The inStates array should be patternCount in length (replication across categoryCount is not required).

Parameters

instance Instance number (input)
tipIndex Index of destination compactBuffer (input)
inStates Pointer to compact states (input)

Returns

error code

BEAGLE_DLLEXPORT int beagleSetTransitionMatrices (int instance, const int * matrixIndices, const double * inMatrices, const double * paddedValues, int count)¶

Set multiple transition matrices. This function copies multiple transition matrices into matrix buffers. This function is used when the application wishes to explicitly set the transition matrices rather than using the beagleSetEigenDecomposition and beagleUpdateTransitionMatrices functions. The inMatrices array should be of size stateCount * stateCount * categoryCount * count.

Parameters

instance Instance number (input)
matrixIndices Indices of matrix buffers (input)
inMatrices Pointer to source transition matrices (input)
paddedValues Values to be used for padding for ambiguous states (e.g. 1 for probability matrices, 0 for derivative matrices) (input)
count Number of transition matrices (input)

Returns

error code

BEAGLE_DLLEXPORT int beagleSetTransitionMatrix (int instance, int matrixIndex, const double * inMatrix, double paddedValue)¶

Set a finite-time transition probability matrix. This function copies a finite-time transition probability matrix into a matrix buffer. This function is used when the application wishes to explicitly set the transition probability matrix rather than using the beagleSetEigenDecomposition and beagleUpdateTransitionMatrices functions. The inMatrix array should be of size stateCount * stateCount * categoryCount and will contain one matrix for each rate category.

Parameters

instance Instance number (input)
matrixIndex Index of matrix buffer (input)
inMatrix Pointer to source transition probability matrix (input)
paddedValue Value to be used for padding for ambiguous states (e.g. 1 for probability matrices, 0 for derivative matrices) (input)

Returns

error code

BEAGLE_DLLEXPORT int beagleUpdatePartials (const int instance, const BeagleOperation * operations, int operationCount, int cumulativeScaleIndex)¶

Calculate or queue for calculation partials using a list of operations. This function either calculates or queues for calculation a list partials. Implementations supporting ASYNCH may queue these calculations while other implementations perform these operations immediately and in order.

Parameters

instance Instance number (input)
operations BeagleOperation list specifying operations (input)
operationCount Number of operations (input)
cumulativeScaleIndex Index number of scaleBuffer to store accumulated factors (input)

Returns

error code

BEAGLE_DLLEXPORT int beagleUpdatePartialsByPartition (const int instance, const BeagleOperationByPartition * operations, int operationCount)¶

Calculate or queue for calculation partials using a list of partition operations. This function either calculates or queues for calculation a list partitioned partials. Implementations supporting ASYNCH may queue these calculations while other implementations perform these operations immediately and in order.

Parameters

instance Instance number (input)
operations BeagleOperation list specifying operations (input)
operationCount Number of operations (input)

Returns

error code

BEAGLE_DLLEXPORT int beagleUpdateTransitionMatrices (int instance, int eigenIndex, const int * probabilityIndices, const int * firstDerivativeIndices, const int * secondDerivativeIndices, const double * edgeLengths, int count)¶

Calculate a list of transition probability matrices. This function calculates a list of transition probabilities matrices and their first and second derivatives (if requested).

Parameters

instance Instance number (input)
eigenIndex Index of eigen-decomposition buffer (input)
probabilityIndices List of indices of transition probability matrices to update (input)
firstDerivativeIndices List of indices of first derivative matrices to update (input, NULL implies no calculation)
secondDerivativeIndices List of indices of second derivative matrices to update (input, NULL implies no calculation)
edgeLengths List of edge lengths with which to perform calculations (input)
count Length of lists

Returns

error code

BEAGLE_DLLEXPORT int beagleUpdateTransitionMatricesWithMultipleModels (int instance, const int * eigenIndices, const int * categoryRateIndices, const int * probabilityIndices, const int * firstDerivativeIndices, const int * secondDerivativeIndices, const double * edgeLengths, int count)¶

Calculate a list of transition probability matrices with multiple models. This function calculates a list of transition probabilities matrices and their first and second derivatives (if requested).

Parameters

instance Instance number (input)
eigenIndices List of indices of eigen-decomposition buffers to use for updates (input)
categoryRateIndices List of indices of category-rate buffers to use for updates (input)
probabilityIndices List of indices of transition probability matrices to update (input)
firstDerivativeIndices List of indices of first derivative matrices to update (input, NULL implies no calculation)
secondDerivativeIndices List of indices of second derivative matrices to update (input, NULL implies no calculation)
edgeLengths List of edge lengths with which to perform calculations (input)
count Length of lists

Returns

error code

BEAGLE_DLLEXPORT int beagleWaitForPartials (const int instance, const int * destinationPartials, int destinationPartialsCount)¶

Block until all calculations that write to the specified partials have completed. This function is optional and only has to be called by clients that 'recycle' partials.

If used, this function must be called after a beagleUpdatePartials call and must refer to indices of 'destinationPartials' that were used in a previous beagleUpdatePartials call. The library will block until those partials have been calculated.

Parameters

instance Instance number (input)
destinationPartials List of the indices of destinationPartials that must be calculated before the function returns
destinationPartialsCount Number of destinationPartials (input)

Returns

error code

Author¶

Generated automatically by Doxygen for BEAGLE from the source code.

Mon Oct 18 2021

Version 3.1.2

Source file:	beagle.h.3.en.gz (from libhmsbeagle-dev 3.1.2+dfsg-13)
Source last updated:	2022-12-16T21:41:31Z
Converted to HTML:	2023-06-10T23:44:44Z