Python wrapper for libffm

This is a python wrapped for LibFFM library writen in C++, orignal implementation is made by Yuchin Juan and Python Wrapped by JRCondeNast.

Features

• [√] Scikit-Learn Compatible API
• [√] Early_Stopping Support
• [√] Evaluation Metric Support
• [√] Pandas Data Converter

Installation

Installing it(Tested in Ubuntu16.04, both Python2 and Python3):

git clone --recursive https://github.com/keyunluo/python-ffm
python3 setup.py install

Usage

import ffm
    
# prepare the data
# (field, index, value) format
    
X = [[(1, 2, 1), (2, 3, 1), (3, 5, 1)],
    [(1, 0, 1), (2, 3, 1), (3, 7, 1)],
    [(1, 1, 1), (2, 3, 1), (3, 7, 1), (3, 9, 1)],
    [(1, 0, 1), (2, 3, 1), (3, 5, 1)], ]
    
y = [1, 1, 0, 1]
    
ffm_data = ffm.FFMData(X, y)
ffm_data_test = ffm.FFMData(X,y)
    
# train the model for 10 iterations 
model = ffm.FFM(eta=0.1, lam=0.0001, k=4)
model.fit(ffm_data,num_iter=10, val_data=ffm_data_test, metric='auc', early_stopping=6, maximum=True) 

# pred proba
print(model.predict_proba(ffm_data_test))
    
# save the model
model.save_model('ololo.bin')
    
# load it to reuse the model
model = ffm.read_model('ololo.bin')

# pred label
print(model.predict(ffm_data_test))

Warning

The package may cause elf library error for some versions of python , please delete the libffm.py file in the installation directory.

Info

This is just a toy, it is recommented to use xLearn for better performance.