def test_svm_predict_convert_dtype(train_dtype, test_dtype, classifier):
X, y = make_classification(n_samples=50, random_state=0)
X = X.astype(train_dtype)
y = y.astype(train_dtype)
X_train, X_test, y_train, y_test = train_test_split(X, y, train_size=0.8,
random_state=0)
if classifier:
clf = cu_svm.SVC()
else:
clf = cu_svm.SVR()
clf.fit(X_train, y_train)
clf.predict(X_test.astype(test_dtype))
def test_svr_skl_cmp(params, dataset, n_rows, n_cols):
""" Compare to Sklearn SVR """
if (dataset == 'Friedman' and n_cols < 5):
# We need at least 5 feature columns for the Friedman dataset
return
X_train, X_test, y_train, y_test = make_regression_dataset(dataset, n_rows,
n_cols)
cuSVR = cu_svm.SVR(**params)
cuSVR.fit(X_train, y_train)
sklSVR = svm.SVR(**params)
sklSVR.fit(X_train, y_train)
compare_svr(cuSVR, sklSVR, X_test, y_test)
def test_svr_skl_cmp_weighted():
""" Compare to Sklearn SVR, use sample weights"""
X, y = make_regression(
n_samples=100, n_features=5, n_informative=2, n_targets=1,
random_state=137, noise=10)
sample_weights = 10*np.sin(np.linspace(0, 2*np.pi, len(y))) + 10.1
params = {'kernel': 'linear', 'C': 10, 'gamma': 1}
cuSVR = cu_svm.SVR(**params)
cuSVR.fit(X, y, sample_weights)
sklSVR = svm.SVR(**params)
sklSVR.fit(X, y, sample_weights)
compare_svr(cuSVR, sklSVR, X, y)
