#!/usr/bin/env python
# Copyright 2016-2019 Biomedical Imaging Group Rotterdam, Departments of
# Medical Informatics and Radiology, Erasmus MC, Rotterdam, The Netherlands
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
from sklearn.base import BaseEstimator
from sklearn.feature_selection import SelectorMixin
import numpy as np
from scipy.stats import ttest_ind, ranksums, mannwhitneyu
[docs]class StatisticalTestThreshold(BaseEstimator, SelectorMixin):
'''
Object to fit feature selection based on statistical tests.
'''
[docs] def __init__(self, metric='ttest', threshold=0.05):
'''
Parameters
----------
metric: string, default 'ttest'
Statistical test used for selection. Options are ttest,
Welch, Wilcoxon, MannWhitneyU
threshold: float, default 0.05
Threshold for p-value in order for feature to be selected
'''
self.metric = metric
self.threshold = threshold
[docs] def fit(self, X_train, Y_train):
'''
Select only features specificed by the metric and threshold per patient.
Parameters
----------
X_train: numpy array, mandatory
Array containing feature values used for model_selection.
Number of objects on first axis, features on second axis.
Y_train: numpy array, mandatory
Array containing the binary labels for each object in X_train.
'''
self.selectrows = list()
self.metric_values = list()
# Set the metric function
if self.metric == 'ttest':
self.metric_function = ttest_ind
self.parameters = {'equal_var': True}
elif self.metric == 'Welch':
self.metric_function = ttest_ind
self.parameters = {'equal_var': False}
elif self.metric == 'Wilcoxon':
self.metric_function = ranksums
self.parameters = {}
elif self.metric == 'MannWhitneyU':
self.metric_function = mannwhitneyu
self.parameters = {}
# Perform the statistical test for each feature
multilabel = type(Y_train[0]) is np.ndarray
for n_feat in range(0, X_train.shape[1]):
# Select only this specific feature for all objects
fv = X_train[:, n_feat]
if multilabel:
# print('Multilabel: take minimum p-value for all label tests.')
# We do a statistical test per label and take the minimum p-value
n_label = Y_train[0].shape[0]
metric_values = list()
for i_label in range(n_label):
class1 = [i for j, i in enumerate(fv) if np.argmax(Y_train[j]) == i_label]
class2 = [i for j, i in enumerate(fv) if np.argmax(Y_train[j]) != i_label]
try:
metric_value_temp = self.metric_function(class1, class2, **self.parameters)[1]
except ValueError as e:
print("[WORC Warning] " + str(e) + '. Replacing metric value by 1.')
metric_value_temp
metric_values.append(metric_value_temp)
metric_value = np.min(metric_values)
else:
# Singlelabel
class1 = [i for j, i in enumerate(fv) if Y_train[j] == 1]
class2 = [i for j, i in enumerate(fv) if Y_train[j] == 0]
try:
metric_value = self.metric_function(class1, class2, **self.parameters)[1]
except ValueError as e:
print("[WORC Warning] " + str(e) + '. Replacing metric value by 1.')
metric_value = 1
self.metric_values.append(metric_value)
if metric_value < self.threshold:
self.selectrows.append(n_feat)
def _get_support_mask(self):
# NOTE: metric is required for the Selector class, but can be empty
pass