def test_get_combined_model(datatype, keys, data_size, fit_intercept, client):
nrows, ncols, n_info = data_size
X_train, y_train, X_test = make_dataset(datatype, nrows, ncols, n_info)
model = LinearRegression(fit_intercept=fit_intercept, client=client)
model.fit(X_train, y_train)
combined_model = model.get_combined_model()
assert combined_model.coef_ is not None
assert combined_model.intercept_ is not None
y_hat = combined_model.predict(X_train.compute())
np.testing.assert_allclose(y_hat.get(),
y_train.compute().get(),
atol=1e-3,
rtol=1e-3)
def test_regressor_mg_train_sg_predict(datatype, keys, data_size,
fit_intercept, client):
nrows, ncols, n_info = data_size
X_train, y_train, X_test = make_dataset(datatype, nrows, ncols, n_info)
X_test_local = X_test.compute()
dist_model = LinearRegression(fit_intercept=fit_intercept, client=client)
dist_model.fit(X_train, y_train)
expected = dist_model.predict(X_test).compute()
local_model = dist_model.get_combined_model()
actual = local_model.predict(X_test_local)
assert_equal(expected.get(), actual.get())
def test_mnmg():
cluster = LocalCUDACluster(threads_per_worker=1)
client = Client(cluster)
n_workers = len(client.scheduler_info()['workers'])
# Create and populate a GPU DataFrame
df_float = cudf.DataFrame()
df_float['0'] = [1.0, 2.0, 5.0]
df_float['1'] = [4.0, 2.0, 1.0]
df_float['2'] = [4., 2, 1]
ddf_float = dask_cudf.from_cudf(df_float, npartitions=2*n_workers)
X = ddf_float[ddf_float.columns.difference(['2'])]
y = ddf_float['2']
mod = LinearRegression()
mod = mod.fit(X, y)
actual_output = str(mod.predict(X).compute().values)
expected_output = '[4. 2. 1.]'
assert actual_output == expected_output