search

baidu / fast_rgf

Multi-core implementation of Regularized Greedy Forest
MIT License
465 stars 111 forks source link

is it possible to release a python wrapper, like xgboost #1

Open fayeshine opened 8 years ago

TongZhang-ML commented 7 years ago

tell me what it is like, and I'd evaluate.

anttttti commented 7 years ago

TZ: other tree ensemble libraries have Python interfaces, both Scikit-learn and non-sklearn. For example, this is a sklearn wrapper for the original RGF library : https://github.com/fukatani/rgf_python

mpearmain commented 7 years ago

+1 great idea for scikit interface. Kaggle community will start using this lib regularly I'm sure

yychenca commented 7 years ago

Creating scikit-learn style wrapper will definitely help the hidden gem RGF to be re-discovered by the data science community.

fukatani commented 6 years ago

Now, rgf_python started to support FastRGF wrapper. We can use FastRGF by scikit-learn-like API.

https://github.com/fukatani/rgf_python/blob/master/FastRGF.rst