Last 7 days
-
r638
(added 4-state optimized CPU version of calcRoot for mixed mo...) committed by daniel.ayres
- added 4-state optimized CPU version of calcRoot for mixed models
added 4-state optimized CPU version of calcRoot for mixed models
-
r637
(fixed bug with second derivative in calcEdge; added empirica...) committed by daniel.ayres
- fixed bug with second derivative in calcEdge; added empirical derivative testing to fourtaxon;
refactored calcRoot in CPU implementation to allow optimized versions (SSE, 4State) of mixed models
fixed bug with second derivative in calcEdge; added empirical derivative testing to fourtaxon;
refactored calcRoot in CPU implementation to allow optimized versions (SSE, 4State) of mixed models
-
r636
(added support for derivative calculation to CPU implementati...) committed by daniel.ayres
- added support for derivative calculation to CPU implementation; added option to fourtaxon to
calculate and print derivatives
added support for derivative calculation to CPU implementation; added option to fourtaxon to
calculate and print derivatives
Last 30 days
-
r635
(reference to platform.h move from libhmsbeagle_la_SOURCES to...) committed by mtholder
- reference to platform.h move from libhmsbeagle_la_SOURCES to library_include_HEADERS. beagle.h
includes platform.h so we have to make sure that platform.h is installed when a user installs the
library (or just inline the content of platform.h into beagle.h)
reference to platform.h move from libhmsbeagle_la_SOURCES to library_include_HEADERS. beagle.h
includes platform.h so we have to make sure that platform.h is installed when a user installs the
library (or just inline the content of platform.h into beagle.h)
-
r634
(Fixed bug in setTransitionMatrix; Ian would like this functi...) committed by msuchard
- Fixed bug in setTransitionMatrix; Ian would like this functionality for porting BEAGLE into xrate
Fixed bug in setTransitionMatrix; Ian would like this functionality for porting BEAGLE into xrate
-
-
r633
(Removed a hard coded preferences for SSE that was overriding...) committed by rambaut
- Removed a hard coded preferences for SSE that was overriding other preferences. Sorry for putting it
there.
Removed a hard coded preferences for SSE that was overriding other preferences. Sorry for putting it
there.
-
r632
(implemented rescaling support for GPU version of calcRootLog...) committed by daniel.ayres
- implemented rescaling support for GPU version of calcRootLogLikelihood when count is > 1
implemented rescaling support for GPU version of calcRootLogLikelihood when count is > 1
-
r631
(implemented GPU version of calcRootLogLikelihood when count ...) committed by daniel.ayres
- implemented GPU version of calcRootLogLikelihood when count is > 1 (though no rescaling support yet)
implemented GPU version of calcRootLogLikelihood when count is > 1 (though no rescaling support yet)
-
r630
(fixed CPU rescaling for calcRootLogLikelihood when count is ...) committed by daniel.ayres
- fixed CPU rescaling for calcRootLogLikelihood when count is > 1
fixed CPU rescaling for calcRootLogLikelihood when count is > 1
-
r629
(added code in tinyTest to test calcRootLogLikelihood when co...) committed by daniel.ayres
- added code in tinyTest to test calcRootLogLikelihood when count is > 1 (works but need to fix
scaling)
added code in tinyTest to test calcRootLogLikelihood when count is > 1 (works but need to fix
scaling)
-
r628
(updated xcode project and removed a couple of minor warnings) committed by daniel.ayres
- updated xcode project and removed a couple of minor warnings
updated xcode project and removed a couple of minor warnings
Earlier this year
-
r627
(Fixed issue with pattern padding in CPU implementations) committed by msuchard
- Fixed issue with pattern padding in CPU implementations
Fixed issue with pattern padding in CPU implementations
-
r626
(In process of debugging pattern padding) committed by dave.swofford
- In process of debugging pattern padding
In process of debugging pattern padding
-
r625
(Changed cpp to hpp in Xcode) committed by msuchard
- Changed cpp to hpp in Xcode
Changed cpp to hpp in Xcode
-
r624
(need to make the pesky config dir
) committed by aaron.darling
- need to make the pesky config dir
need to make the pesky config dir
-
r623
(simplify optimization flags for profile-guided optimization,...) committed by aaron.darling
- simplify optimization flags for profile-guided optimization, reduce requirement to sse2
simplify optimization flags for profile-guided optimization, reduce requirement to sse2
-
r622
(SSE now returns the SSE flag in its instance description.) committed by rambaut
- SSE now returns the SSE flag in its instance description.
SSE now returns the SSE flag in its instance description.
-
r621
(Added a few more beagle options to BEAST main.) committed by rambaut
- Added a few more beagle options to BEAST main.
Added a few more beagle options to BEAST main.
-
r620
(Possible fix for bug on GTX295) committed by msuchard
- Possible fix for bug on GTX295
Possible fix for bug on GTX295
-
r619
(Started to convert 4StateSSEImpl into a template to easily a...) committed by msuchard
- Started to convert 4StateSSEImpl into a template to easily allow for a singe-precision version
Started to convert 4StateSSEImpl into a template to easily allow for a singe-precision version
-
r618
(Fixed getPartials and getTransitionMatrix to handle padding ...) committed by msuchard
- Fixed getPartials and getTransitionMatrix to handle padding (always for matrices and sometimes for
partials)
Fixed getPartials and getTransitionMatrix to handle padding (always for matrices and sometimes for
partials)
-
r617
(Stage 2 of adding beagleGetTransitionMatrix to JNI interface) committed by msuchard
- Stage 2 of adding beagleGetTransitionMatrix to JNI interface
Stage 2 of adding beagleGetTransitionMatrix to JNI interface
-
r616
(Stage 1 is adding beagleGetTransitionMatrix to the API) committed by msuchard
- Stage 1 is adding beagleGetTransitionMatrix to the API
Stage 1 is adding beagleGetTransitionMatrix to the API
-
r615
(Added pattern padding in CPU implementations for vectorizati...) committed by msuchard
- Added pattern padding in CPU implementations for vectorization across patterns
Added pattern padding in CPU implementations for vectorization across patterns
-
r614
(Removed duplicated implementation names) committed by msuchard
- Removed duplicated implementation names
Removed duplicated implementation names
-
r613
(Added some more BEAGLE options to the main dialog box.) committed by rambaut
- Added some more BEAGLE options to the main dialog box.
Added some more BEAGLE options to the main dialog box.
-
r612
(Added some BEAST command-line options to control the BEAGLE ...) committed by rambaut
- Added some BEAST command-line options to control the BEAGLE preference flags. For example
-beagle_single to ask for single precision. These actually set a system preference
beagle.preferred.flags so this can be set by the JVM. Cleaned up the presentation of information
about what resource is being used - now the 'instance flags' are shown to show info about the
implementation actually being used.
Added some BEAST command-line options to control the BEAGLE preference flags. For example
-beagle_single to ask for single precision. These actually set a system preference
beagle.preferred.flags so this can be set by the JVM. Cleaned up the presentation of information
about what resource is being used - now the 'instance flags' are shown to show info about the
implementation actually being used.
-
r611
(trying inline to avoid duplicate symbols in link
) committed by aaron.darling
- trying inline to avoid duplicate symbols in link
trying inline to avoid duplicate symbols in link
-
r610
(Refactored MEMCNV to beagleMemCpy to avoid conflict with mac...) committed by msuchard
- Refactored MEMCNV to beagleMemCpy to avoid conflict with macro defined in GPU
Refactored MEMCNV to beagleMemCpy to avoid conflict with macro defined in GPU
-
r609
(adding OpenMP for generic likelihood calc
) committed by aaron.darling
- adding OpenMP for generic likelihood calc
adding OpenMP for generic likelihood calc
-
r608
(adding vis studio installer project to make dist
) committed by aaron.darling
- adding vis studio installer project to make dist
adding vis studio installer project to make dist
-
r607
(removing extraneous #include
) committed by aaron.darling
- removing extraneous #include
removing extraneous #include
-
r606
(removing extraneous #include
) committed by aaron.darling
- removing extraneous #include
removing extraneous #include
-
r605
(templating MEMCNV to eliminate some of the if(DOUBLE_PRECISI...) committed by aaron.darling
- templating MEMCNV to eliminate some of the if(DOUBLE_PRECISION) statements
templating MEMCNV to eliminate some of the if(DOUBLE_PRECISION) statements
-
r604
(Implement a different template-based naming function
) committed by aaron.darling
- Implement a different template-based naming function
Implement a different template-based naming function
-
r603
(fix for dependency tracking, at least on linux
) committed by aaron.darling
- fix for dependency tracking, at least on linux
fix for dependency tracking, at least on linux
-
r602
(Implemented getPartials with rescaling for the CPU) committed by msuchard
- Implemented getPartials with rescaling for the CPU
Implemented getPartials with rescaling for the CPU
-
r601
(Some not-so-good compilation instructions; as INSTALL does n...) committed by msuchard
- Some not-so-good compilation instructions; as INSTALL does not mention autogen.sh
Some not-so-good compilation instructions; as INSTALL does not mention autogen.sh
-
r600
(give appropriate visibility to base-class members
) committed by aaron.darling
- give appropriate visibility to base-class members
give appropriate visibility to base-class members
-
r599
(Reduced memory footprint of GPU-complex implementation) committed by msuchard
- Reduced memory footprint of GPU-complex implementation
Reduced memory footprint of GPU-complex implementation
-
r598
(Removing unnecessary kernels and related code) committed by msuchard
- Removing unnecessary kernels and related code
Removing unnecessary kernels and related code
-
r597
(Complex GPU algorithm now works for stateCount == 4 under bo...) committed by msuchard
- Complex GPU algorithm now works for stateCount == 4 under both MULTIPLY_BLOCK_SIZE = 16 and = 2, so
it should work for any arbitrary stateCount as well
Complex GPU algorithm now works for stateCount == 4 under both MULTIPLY_BLOCK_SIZE = 16 and = 2, so
it should work for any arbitrary stateCount as well
-
r596
(don't change the PATH env variable because MSI installer wil...) committed by aaron.darling
- don't change the PATH env variable because MSI installer will then delete it on uninstall!
don't change the PATH env variable because MSI installer will then delete it on uninstall!
-
r595
(fix for wrong project name in installer build) committed by aaron.darling
- fix for wrong project name in installer build
fix for wrong project name in installer build
-
r594
(New algorithm for complex GPU calculations; should be more s...) committed by msuchard
- New algorithm for complex GPU calculations; should be more scalabe for large stateCounts
New algorithm for complex GPU calculations; should be more scalabe for large stateCounts
-
r593
(adding a script that will build all targets and create an in...) committed by aaron.darling
- adding a script that will build all targets and create an installer using viz studio 2008
adding a script that will build all targets and create an installer using viz studio 2008
-
r592
(adding a windows installer project for viz studio 2008. cou...) committed by aaron.darling
- adding a windows installer project for viz studio 2008. could use some testing.
adding a windows installer project for viz studio 2008. could use some testing.
-
r591
(Complex GPU works for stateCount == 4; further work is neede...) committed by msuchard
- Complex GPU works for stateCount == 4; further work is needed general stateCounts
Complex GPU works for stateCount == 4; further work is needed general stateCounts
-
r590
(Added R script to generate test values for complex diagonali...) committed by msuchard
- Added R script to generate test values for complex diagonalization
Added R script to generate test values for complex diagonalization
|