mrmr_fcq.py

import pandas as pd
from sklearn.feature_selection import f_regression

# inputs:
#    X: pandas.DataFrame, features
#    y: pandas.Series, target variable
#    K: number of features to select

# compute F-statistics and correlations
F = pd.Series(f_regression(X, y)[0], index = X.columns)
corr = X.corr().abs().clip(.00001) # minimum value of correlation set to .00001 (to avoid division by zero)

# initialize list of selected features and list of excluded features
selected = []
not_selected = list(X.columns)

# repeat K times: 
# compute FCQ score for all the features that are currently excluded,
# then find the best one, add it to selected, and remove it from not_selected
for i in range(K):
    
    # compute FCQ score for all the (currently) excluded features (this is Formula 2)
    score = F.loc[not_selected] / corr.loc[not_selected, selected].mean(axis = 1).fillna(.00001)
    
    # find best feature, add it to selected and remove it from not_selected
    best = score.index[score.argmax()]
    selected.append(best)
    not_selected.remove(best)