def test_get_dtype_from_model_func():
X, y = make_regression(n_samples=81, n_features=10, noise=0.1,
random_state=42, dtype=np.float32)
# checking model with float32 dtype
model_f32 = reg().fit(X, y)
assert get_dtype_from_model_func(model_f32.predict) == np.float32
# checking model with float64 dtype
X = X.astype(np.float64)
y = y.astype(np.float64)
model_f64 = reg().fit(X, y)
assert get_dtype_from_model_func(model_f64.predict) == np.float64
# checking model that has not been fitted yet
model_not_fit = reg()
assert(get_dtype_from_model_func(model_not_fit.predict) is None)
# checking arbitrary function
def dummy_func(x):
return x + x
assert get_dtype_from_model_func(dummy_func) is None
def test_model_func_call_gpu():
X, y = make_regression(n_samples=81, n_features=10, noise=0.1,
random_state=42, dtype=np.float32)
model = reg().fit(X, y)
z = model_func_call(X=X,
model_func=model.predict,
gpu_model=True)
assert isinstance(z, cp.ndarray)
z = model_func_call(X=cp.asnumpy(X),
model_func=dummy_func,
gpu_model=False)
assert isinstance(z, cp.ndarray)
with pytest.raises(TypeError):
z = model_func_call(X=X,
model_func=dummy_func,
gpu_model=True)
model = PCA(n_components=10).fit(X)
z = model_func_call(X=X,
model_func=model.transform,
gpu_model=True)
assert isinstance(z, cp.ndarray)
def make_regression_dataset(datatype, nrows, ncols):
ninformative = max(min(ncols, 5), int(math.ceil(ncols / 5)))
X, y = data.make_regression(dtype=datatype,
n_samples=nrows + 1000,
n_features=ncols,
random_state=SEED,
n_informative=ninformative)
return dsel.train_test_split(X, y, random_state=SEED, train_size=nrows)
def test_typeerror_input():
X, y = make_regression(n_samples=100, n_features=10, random_state=10)
clf = Lasso()
clf.fit(X, y)
exp = KernelExplainer(model=clf.predict, data=X, nsamples=10)
try:
_ = exp.shap_values(X)
assert True
except TypeError:
assert False
def make_dataset(request):
nrows, ncols, n_info, datatype = request.param
X, y = make_regression(n_samples=nrows,
n_informative=n_info,
n_features=ncols,
random_state=0)
X = cp.array(X).astype(datatype)
y = cp.array(y).astype(datatype)
X_train, X_test, y_train, y_test = train_test_split(X,
y,
train_size=0.8,
random_state=10)
return nrows, datatype, X_train, X_test, y_train, y_test
def make_dataset(request):
nrows, ncols, n_info, datatype = request.param
if nrows == 500000 and datatype == np.float64 and \
pytest.max_gpu_memory < 32:
if pytest.adapt_stress_test:
nrows = nrows * pytest.max_gpu_memory // 32
else:
pytest.skip("Insufficient GPU memory for this test."
"Re-run with 'CUML_ADAPT_STRESS_TESTS=True'")
X, y = make_regression(n_samples=nrows,
n_informative=n_info,
n_features=ncols,
random_state=0)
X = cp.array(X).astype(datatype)
y = cp.array(y).astype(datatype)
X_train, X_test, y_train, y_test = train_test_split(X,
y,
train_size=0.8,
random_state=10)
return nrows, datatype, X_train, X_test, y_train, y_test
def regression_dataset(request):
X, y = make_regression(n_samples=10, n_features=5, random_state=0)
return train_test_split(X, y, random_state=0)