scikit-learn-fastLDA

An optimized cython version of LDA model implemented in scikit-learn

Installation

I just show the procedure for a Mac using pyenv (feel free to use your python library and environment manager):

brew install readline xz
brew install pyenv pyenv-virtualenv

pyenv install 3.7.9
pyenv virtualenv 3.7.9 fastLDA_3.7

cd {YOUR_DIR_PATH}/scikit-learn-LDAfast
pyenv local fastLDA_3.7

Now open a new terminal and navigate in the code directory. You should have "fastLDA" environment active. If so please launch the following commands

pip install -r requirements.txt
python -m ipykernel install --user --name fastLDA_3.7 --display-name "fast LDA  3.7"

Now it is time to compile your Cythonic code:

cd LDAfast/clib
python setup.py build_ext --inplace

Well done, you are ready to use the fastLDA method!

API

See the example notebook on how you can call the fastLDA method and the scikit-learn documentation for the use, args and the parameters definition.

Good decomposition!