def test_Ridge(self):
breast_cancer = load_breast_cancer()
X = breast_cancer.data
y = breast_cancer.target
X_train, X_test, y_train, y_test = train_test_split(X,
y,
test_size=0.2,
random_state=71023)
wine = load_wine()
Z = wine.data
t = wine.target
Z_train, Z_test, t_train, t_test = train_test_split(Z,
t,
test_size=0.2,
random_state=61283)
fit_obj = ns.Ridge2Classifier(lambda1=0.025,
lambda2=0.5,
n_hidden_features=5,
n_clusters=0)
fit_obj2 = ns.Ridge2Classifier(lambda1=0.01,
lambda2=0.01,
n_hidden_features=10,
n_clusters=2)
fit_obj3 = ns.Ridge2Classifier(lambda1=0.025,
lambda2=0.05,
n_hidden_features=5,
n_clusters=0)
fit_obj4 = ns.Ridge2Classifier(lambda1=0.001,
lambda2=0.01,
n_hidden_features=10,
n_clusters=2)
fit_obj.fit(X_train, y_train)
preds1 = fit_obj.predict_proba(X_test)
fit_obj2.fit(X_train, y_train)
preds2 = fit_obj.predict_proba(X_test)
fit_obj3.fit(Z_train, t_train)
preds3 = fit_obj3.predict_proba(Z_test)
fit_obj4.fit(Z_train, t_train)
preds4 = fit_obj4.predict_proba(Z_test)
self.assertTrue(np.allclose(preds1[0, 0], 5.1412488698250085e-06))
self.assertTrue(np.allclose(preds2[0, 0], 5.1412488698250085e-06))
self.assertTrue(np.allclose(preds3[0, 0], 0.0488412175))
self.assertTrue(np.allclose(preds4[0, 0], 0.8545733738))
self.assertTrue(
np.allclose(fit_obj.predict(X_test)[0], 1)
& np.allclose(fit_obj2.predict(X_test)[0], 1)
& np.allclose(fit_obj3.predict(Z_test)[0], 1)
& np.allclose(fit_obj4.predict(Z_test)[0], 0))
def test_score(self):
breast_cancer = load_breast_cancer()
X = breast_cancer.data
y = breast_cancer.target
np.random.seed(123)
X_train, X_test, y_train, y_test = train_test_split(X,
y,
test_size=0.2)
wine = load_wine()
Z = wine.data
t = wine.target
np.random.seed(123)
Z_train, Z_test, t_train, t_test = train_test_split(Z,
t,
test_size=0.2)
fit_obj = ns.Ridge2Classifier(lambda1=0.025,
lambda2=0.5,
n_hidden_features=5,
n_clusters=2)
fit_obj2 = ns.Ridge2Classifier(lambda1=0.01,
lambda2=0.01,
n_hidden_features=10,
n_clusters=2)
fit_obj3 = ns.Ridge2Classifier(lambda1=0.025,
lambda2=0.05,
n_hidden_features=5,
n_clusters=2)
fit_obj4 = ns.Ridge2Classifier(lambda1=0.001,
lambda2=0.01,
n_hidden_features=10,
n_clusters=2)
fit_obj.fit(X_train, y_train)
score1 = fit_obj.score(X_test, y_test)
fit_obj2.fit(X_train, y_train)
score2 = fit_obj2.score(X_test, y_test)
fit_obj3.fit(Z_train, t_train)
score3 = fit_obj3.score(Z_test, t_test)
fit_obj4.fit(Z_train, t_train)
score4 = fit_obj4.score(Z_test, t_test)
self.assertTrue(np.allclose(score1, 0.9649122807017544))
self.assertTrue(np.allclose(score2, 0.94736842105263153))
self.assertTrue(np.allclose(score3, 1.0))
self.assertTrue(np.allclose(score4, 0.9722222222222222))
# dataset no. 1 ----------
breast_cancer = load_breast_cancer()
X = breast_cancer.data
y = breast_cancer.target
# split data into training test and test set
np.random.seed(123)
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)
# create the model with nnetsauce
fit_obj = ns.Ridge2Classifier(lambda1=6.90185578e+04,
lambda2=3.17392781e+02,
n_hidden_features=95,
n_clusters=2,
row_sample=4.63427734e-01,
dropout=3.62817383e-01,
type_clust="gmm")
# fit the model on training set
fit_obj.fit(X_train, y_train)
# get the accuracy on test set
print(fit_obj.score(X_test, y_test))
# get area under the curve on test set (auc)
print(fit_obj.score(X_test, y_test, scoring="roc_auc"))
# dataset no. 2 ----------