def test_SklearnWrapper_fit_predict_with_clipped_input_data():
nz = 5
model = _RegressorEnsemble(
base_regressor=DummyRegressor(strategy="constant",
constant=np.arange(nz)),
n_jobs=1,
)
wrapper = SklearnWrapper(
input_variables=["a", "b"],
output_variables=["c"],
model=model,
packer_config=PackerConfig({"a": {
"z": SliceConfig(2, None)
}}),
)
dims = ["x", "y", "z"]
shape = (2, 2, nz)
arr = np.arange(np.prod(shape)).reshape(shape)
input_data = xr.Dataset({
"a": (dims, arr),
"b": (dims[:-1], arr[:, :, 0]),
"c": (dims, arr + 1)
})
wrapper.fit([input_data])
wrapper.predict(input_data)
def test_SklearnWrapper_serialize_fit_after_load(tmpdir):
model = _RegressorEnsemble(base_regressor=LinearRegression(), n_jobs=1)
wrapper = SklearnWrapper(
input_variables=["x"],
output_variables=["y"],
model=model,
)
# setup input data
dims = ["unstacked_dim", "z"]
data = xr.Dataset({
"x": (dims, np.ones((1, 1))),
"y": (dims, np.ones((1, 1)))
})
wrapper.fit([data])
# serialize/deserialize
path = str(tmpdir)
wrapper.dump(path)
# fit loaded model
loaded = wrapper.load(path)
loaded.fit([data])
assert len(loaded.model.regressors) == 2
def test_SklearnWrapper_serialize_predicts_the_same(tmpdir, scale_factor):
# Setup wrapper
if scale_factor:
scaler = ManualScaler(np.array([scale_factor]))
else:
scaler = None
model = _RegressorEnsemble(base_regressor=LinearRegression(), n_jobs=1)
wrapper = SklearnWrapper(
input_variables=["x"],
output_variables=["y"],
model=model,
)
wrapper.target_scaler = scaler
# setup input data
dims = ["unstacked_dim", "z"]
data = xr.Dataset({
"x": (dims, np.ones((1, 1))),
"y": (dims, np.ones((1, 1)))
})
wrapper.fit([data])
# serialize/deserialize
path = str(tmpdir)
wrapper.dump(path)
loaded = wrapper.load(path)
xr.testing.assert_equal(loaded.predict(data), wrapper.predict(data))
def test_regressor_ensemble():
base_regressor = LinearRegression()
ensemble_regressor = _RegressorEnsemble(base_regressor, n_jobs=1)
num_batches = 3
X = np.array([[1, 1], [1, 2], [2, 2], [2, 3]])
y = np.dot(X, np.array([1, 2])) + 3
for i in range(num_batches):
ensemble_regressor.fit(X, y)
return ensemble_regressor
def test_SklearnWrapper_raises_not_implemented_error_with_clipped_output_data(
):
nz = 5
model = _RegressorEnsemble(
base_regressor=DummyRegressor(strategy="constant",
constant=np.arange(nz)),
n_jobs=1,
)
with pytest.raises(NotImplementedError):
SklearnWrapper(
input_variables=["a", "b"],
output_variables=["c"],
model=model,
packer_config=PackerConfig({"c": {
"z": SliceConfig(2, None)
}}),
)
def fit_wrapper_with_columnar_data():
nz = 2
model = _RegressorEnsemble(
base_regressor=DummyRegressor(strategy="constant",
constant=np.arange(nz)),
n_jobs=1,
)
wrapper = SklearnWrapper(
input_variables=["a"],
output_variables=["b"],
model=model,
)
dims = ["x", "y", "z"]
shape = (2, 2, nz)
arr = np.arange(np.prod(shape)).reshape(shape)
input_data = xr.Dataset({"a": (dims, arr), "b": (dims, arr + 1)})
wrapper.fit([input_data])
return input_data, wrapper