Skip to content
Snippets Groups Projects

Repository graph

You can move around the graph by using the arrow keys.
Select Git revision
No matching results
An error occurred while fetching branches. Retry the search.
An error occurred while fetching tags. Retry the search.
Created with Raphaël 2.2.026Nov10825Oct24211312118528Sep25203227Aug26252423221716141098765432131Jul3029282726251615141312111097542125Jun242019181714Fixed AdEllpack up to 99x99 matrices. Larger matrices have increasingly different resulting vectors. FIXME.Removed useless comments.Commit for backup purposes of debugging AdELL.Changed cross-device assignment for AdEll (broken in certain scenarios).Removed commented out code. Finalized cross-device assignment (Not ideal for CPU, but enough for benchmarking GPU).Removed commented out code.Removed HostBadAlloc. Uncommented benchmarks. Commented prints.Disabled part of the benchmark for debugging purposes.Added basic functionality for cross-device copy assignment. Removed StripSize template typename as it was never used anywhere.Temporarily disabled operator= tests for ChunkedEllpack and enabled them for BiEllpack.Changed tests for BiELL.Fixed AdEllpack friend class.Changed setElement test to a more complex matrix.Fixed BiEllpack. Uncommented tests that didn't work before. Removed useless comments.Fixed setCompressedRowLengths by getting around const vector. Fixed getStripLength and getGroupLength by using accessing methods instead of []. Fixed addElement to multiply the present value rather than the value that is being added.Removed useless comments.Added operator functions (some not done). Removed useless comments.Added comparison and assignment operators.Removed useless comments.Removed useless comments.Removed useless comments.Fixed allocation of elements, empty matrices don't throw errors nowRemoved prints. Comitting for backup purposesImplemented fix for case when roundToMultiple would give a number smaller than the number of rows, thus causing index to be out of bounds on CUDA. Debugging prints includedImplemented fix for case when roundToMultiple would give a number smaller than the number of rows, thus causing index to be out of bounds on CUDAImplemented fix for case when roundToMultiple would give a number smaller than the number of rows, thus causing index to be out of bounds on CUDAAdded checking for negative number of elements. Debuggin prints includedAdded checking for negative number of elementsAdded checking for negative number of elementsAdded debuggin printsAdded checking for negative number of elements, in case of int overflow. Debugging prints included.Debugging Ellpack errorsInitial commitAdded exception handling for allocation on CPUUncommented all tests. Updated comments.Removed print outs for checking validity of operator overloading.Changed default precision from all to double.Uncommented result comparison. Changed benchmark to only benchmark Chunked Ellpack.Implemented ChunkedEllpack cross-device assignment, with error checking prints.Added Vector types to help with operator= overloading for cross device assignment.