Name		Name	Last commit message	Last commit date
wikisim_server		wikisim_server
.gitignore		.gitignore
LICENSE		LICENSE
README.md		README.md
plots.py		plots.py
prepare_shootout.py		prepare_shootout.py
requirements.txt		requirements.txt
run_all.sh		run_all.sh
shootout.py		shootout.py
shootout_annoy.py		shootout_annoy.py
shootout_flann.py		shootout_flann.py
shootout_gensim.py		shootout_gensim.py
shootout_kgraph.py		shootout_kgraph.py
shootout_lsh.py		shootout_lsh.py
shootout_sklearn.py		shootout_sklearn.py

Repository files navigation

sim-shootout

Code for the blog series on benchmarking nearest-neighbour libraries:

Part II: Contestants

Part III: Querying

The target are libraries that expose a Python interface and allow (exact or approximate) k-NN search in high dimensional vector spaces.

Run run_all.sh to download Wikipedia, index it and run all experiments.

About

Code for "Performance shootout between nearest-neighbour libraries": http://radimrehurek.com/2013/11/performance-shootout-of-nearest-neighbours-intro

98 stars

9 watching

Report repository

Releases

Packages

Used by

Contributors

Languages