def test_plot_roc(self): M, labels = uft.generate_correlated_test_matrix(1000) M_train, M_test, labels_train, labels_test = train_test_split( M, labels) clf = RandomForestClassifier(random_state=0) clf.fit(M_train, labels_train) score = clf.predict_proba(M_test)[:, -1] fig = comm.plot_roc(labels_test, score, verbose=False) self.add_fig_to_report(fig, 'plot_roc')
def test_plot_roc(self): M, labels = uft.generate_correlated_test_matrix(1000) M_train, M_test, labels_train, labels_test = train_test_split( M, labels) clf = RandomForestClassifier(random_state=0) clf.fit(M_train, labels_train) score = clf.predict_proba(M_test)[:,-1] fig = dsp.plot_roc(labels_test, score, verbose=False) self.add_fig_to_report(fig, 'plot_roc')
def test_get_roc_auc(self): M, labels = uft.generate_correlated_test_matrix(1000) M_train, M_test, labels_train, labels_test = train_test_split( M, labels) clf = RandomForestClassifier(random_state=0) clf.fit(M_train, labels_train) score = clf.predict_proba(M_test)[:, -1] self.assertTrue( np.allclose(comm.get_roc_auc(labels_test, score, verbose=False), roc_auc_score(labels_test, score)))
def test_get_roc_auc(self): M, labels = uft.generate_correlated_test_matrix(1000) M_train, M_test, labels_train, labels_test = train_test_split( M, labels) clf = RandomForestClassifier(random_state=0) clf.fit(M_train, labels_train) score = clf.predict_proba(M_test)[:,-1] self.assertTrue(np.allclose( dsp.get_roc_auc(labels_test, score, verbose=False), roc_auc_score(labels_test, score)))