def cife(X, y, **kwargs):
"""
This function implements the CIFE feature selection
Input
-----
X: {numpy array}, shape (n_samples, n_features)
input data, guaranteed to be discrete
y: {numpy array}, shape (n_samples,)
input class labels
kwargs: {dictionary}
n_selected_features: {int}
number of features to select
Output
------
F: {numpy array}, shape (n_features,)
index of selected features, F[0] is the most important feature
J_CMI: {numpy array}, shape: (n_features,)
corresponding objective function value of selected features
MIfy: {numpy array}, shape: (n_features,)
corresponding mutual information between selected features and response
Reference
---------
Brown, Gavin et al. "Conditional Likelihood Maximisation: A Unifying Framework for Information Theoretic Feature Selection." JMLR 2012.
"""
if 'n_selected_features' in kwargs.keys():
n_selected_features = kwargs['n_selected_features']
dictOfHeader = kwargs['dict_features']
F, J_CMI, MIfy, dictFeatJcmi = LCSI.lcsi(X, y, beta=1, gamma=1, n_selected_features=n_selected_features, dict_features=dictOfHeader)
else:
F, J_CMI, MIfy, dictFeatJcmi = LCSI.lcsi(X, y, beta=1, gamma=1)
return F, J_CMI, MIfy, dictFeatJcmi
def mrmr(X, y, **kwargs):
"""
This function implements the MRMR feature selection
Input
-----
X: {numpy array}, shape (n_samples, n_features)
input data, guaranteed to be discrete
y: {numpy array}, shape (n_samples,)
input class labels
kwargs: {dictionary}
n_selected_features: {int}
number of features to select
Output
------
F: {numpy array}, shape (n_features,)
index of selected features, F[0] is the most important feature
J_CMI: {numpy array}, shape: (n_features,)
corresponding objective function value of selected features
MIfy: {numpy array}, shape: (n_features,)
corresponding mutual information between selected features and response
Reference
---------
Brown, Gavin et al. "Conditional Likelihood Maximisation: A Unifying Framework for Information Theoretic Feature Selection." JMLR 2012.
"""
if 'n_selected_features' in kwargs.keys():
n_selected_features = kwargs['n_selected_features']
F, J_CMI, MIfy = LCSI.lcsi(X, y, gamma=0, function_name='MRMR', n_selected_features=n_selected_features)
else:
F, J_CMI, MIfy = LCSI.lcsi(X, y, gamma=0, function_name='MRMR')
return F, J_CMI, MIfy
def mim(X, y, **kwargs):
"""
This function implements the MIM feature selection
Input
-----
X: {numpy array}, shape (n_samples, n_features)
input data, guaranteed to be discrete
y: {numpy array}, shape (n_samples,)
input class labels
kwargs: {dictionary}
n_selected_features: {int}
number of features to select
Output
------
F: {numpy array}, shape (n_features, )
index of selected features, F[0] is the most important feature
Reference
---------
Brown, Gavin et al. "Conditional Likelihood Maximisation: A Unifying Framework for Information Theoretic Feature Selection." JMLR 2012.
"""
if 'n_selected_features' in kwargs.keys():
n_selected_features = kwargs['n_selected_features']
F = LCSI.lcsi(X,
y,
beta=0,
gamma=0,
n_selected_features=n_selected_features)
else:
F = LCSI.lcsi(X, y, beta=0, gamma=0)
return F
def mifs(X, y, mode="rank", **kwargs):
"""
This function implements the MIFS feature selection
Input
-----
X: {numpy array}, shape (n_samples, n_features)
input data, guaranteed to be discrete
y: {numpy array}, shape (n_samples,)
input class labels
kwargs: {dictionary}
n_selected_features: {int}
number of features to select
Output
------
F: {numpy array}, shape (n_features,)
index of selected features, F[0] is the most important feature
J_CMI: {numpy array}, shape: (n_features,)
corresponding objective function value of selected features
MIfy: {numpy array}, shape: (n_features,)
corresponding mutual information between selected features and response
Reference
---------
Brown, Gavin et al. "Conditional Likelihood Maximisation: A Unifying Framework for Information Theoretic Feature Selection." JMLR 2012.
"""
if 'beta' not in list(kwargs.keys()):
beta = 0.5
else:
beta = kwargs['beta']
if 'n_selected_features' in list(kwargs.keys()):
n_selected_features = kwargs['n_selected_features']
F, J_CMI, MIfy = LCSI.lcsi(X,
y,
beta=beta,
gamma=0,
n_selected_features=n_selected_features)
else:
F, J_CMI, MIfy = LCSI.lcsi(X, y, beta=beta, gamma=0)
if mode == "index":
return np.array(F, dtype=int)
else:
# make sure that F is the same size??
return reverse_argsort(F, size=X.shape[1])
# if su_calculation(fp, fi) >= t1[i, 1]:
# # construct the mask for feature whose su is larger than su(fp,y)
# idx = s_list[:, 0] != i
# idx = np.array([idx, idx])
# idx = np.transpose(idx)
# # delete the feature by using the mask
# s_list = s_list[idx]
# length = len(s_list)//2
# s_list = s_list.reshape((length, 2))
# return np.array(F, dtype=int), np.array(SU)
#
#feat_index, sym_arr = fcbf(X_train_data.iloc[:,:5], X_test_data.iloc[:,:5])
#MIM
from skfeature.function.information_theoretical_based import LCSI
F, J_CMI, MIfy = LCSI.lcsi(X_train_data, y_train_data, beta=0, gamma=0)
from sklearn.ensemble import AdaBoostRegressor
regr = AdaBoostRegressor(random_state=0, n_estimators=100)
regr.fit(X_train_data, y_train_data)
imp = regr.feature_importances_
X_train_data.columns[imp > 0]
from chefboost import Chefboost as chef
import pandas as pd
config = {'algorithm': 'C4.5'}
df = X_train_data
df["Decision"] = y_train_data