Skip to content

HTTPS clone URL

Subversion checkout URL

You can clone with HTTPS or Subversion.

Download ZIP
Benchmark of different sparse-sparse and sparse-dense dot product implementations.
C++ Python
branch: master

Fetching latest commit…

Cannot retrieve the latest commit at this time

Failed to load latest commit information.
Makefile
README.rst
bench.py
dot.cpp
dotbench.pyx
setup.py

README.rst

dot-bench

This is an ongoing project which aims to benchmark different implementations for sparse-sparse and sparse-dense dot products, especially in the context of Machine Learning algorithms.

  • Python: Generate data, plots and reports.
  • C/C++: Perform computations.
  • Cython: Wrap C code.

Note: Cython is not used for computations to allow more people to read the code (the benchmark is intended for a broad audience, not only Python programmers). Moreover, handwritten C may be more efficient.

ToDO

  • Plots and reports for different matrix and weight sparsities
  • Merge, B-tree, Hash-table (google-sparse-hash?) based sparse-sparse implementations
  • Comparison of different scenarios: training vs prediction, real vs binary features
Something went wrong with that request. Please try again.