def test_getattr(client):
# Test getattr on local param
kmeans_model = KMeans(client=client)
# Test AttributeError
with pytest.raises(AttributeError):
kmeans_model.cluster_centers_
assert kmeans_model.client is not None
# Test getattr on local_model param with a non-distributed model
X, y = make_blobs(n_samples=5,
n_features=5,
centers=2,
n_parts=2,
cluster_std=0.01,
random_state=10)
kmeans_model.fit(X)
assert kmeans_model.cluster_centers_ is not None
assert isinstance(kmeans_model.cluster_centers_, cupy.ndarray)
# Test getattr on trained distributed model
X, y = load_text_corpus(client)
nb_model = MultinomialNB(client=client)
nb_model.fit(X, y)
assert nb_model.feature_count_ is not None
assert isinstance(nb_model.feature_count_, cupy.ndarray)
def fit(self, X_train, y_train):
""" Fit decision tree model """
if 'XGBoost' in self.hpo_config.model_type:
hpo_log.info('> fit xgboost model')
dtrain = xgboost.dask.DaskDMatrix(self.client, X_train, y_train)
num_boost_round = self.hpo_config.model_params['num_boost_round']
xgboost_output = xgboost.dask.train(
self.client,
self.hpo_config.model_params,
dtrain,
num_boost_round=num_boost_round)
trained_model = xgboost_output['booster']
elif 'RandomForest' in self.hpo_config.model_type:
hpo_log.info('> fit randomforest model')
trained_model = RandomForestClassifier(
n_estimators=self.hpo_config.model_params['n_estimators'],
max_depth=self.hpo_config.model_params['max_depth'],
max_features=self.hpo_config.model_params['max_features'],
n_bins=self.hpo_config.model_params['n_bins']).fit(
X_train, y_train.astype('int32'))
elif 'KMeans' in self.hpo_config.model_type:
hpo_log.info('> fit kmeans model')
trained_model = KMeans(
n_clusters=self.hpo_config.model_params['n_clusters'],
max_iter=self.hpo_config.model_params['max_iter'],
random_state=self.hpo_config.model_params['random_state'],
init=self.hpo_config.model_params['init']).fit(X_train)
return trained_model
def test_getattr(cluster):
client = Client(cluster)
try:
# Test getattr on local param
kmeans_model = KMeans(client=client)
assert kmeans_model.client is not None
# Test getattr on local_model param with a non-distributed model
X, y = make_blobs(n_samples=5,
n_features=5,
centers=2,
n_parts=2,
cluster_std=0.01,
random_state=10)
wait(X)
kmeans_model.fit(X)
assert kmeans_model.cluster_centers_ is not None
assert isinstance(kmeans_model.cluster_centers_, cupy.core.ndarray)
# Test getattr on trained distributed model
X, y = load_text_corpus(client)
print(str(X.compute()))
nb_model = MultinomialNB(client=client)
nb_model.fit(X, y)
assert nb_model.feature_count_ is not None
assert isinstance(nb_model.feature_count_, cupy.core.ndarray)
finally:
client.close()
def test_getattr(cluster):
client = Client(cluster)
# Test getattr on local param
kmeans_model = KMeans(client=client)
assert kmeans_model.client is not None
# Test getattr on local_model param with a non-distributed model
X_cudf, y = make_blobs(5,
5,
2,
2,
cluster_std=0.01,
verbose=False,
random_state=10)
wait(X_cudf)
kmeans_model.fit(X_cudf)
assert kmeans_model.cluster_centers_ is not None
assert isinstance(kmeans_model.cluster_centers_, cudf.DataFrame)
# Test getattr on trained distributed model
X, y = load_text_corpus(client)
print(str(X.compute()))
nb_model = MultinomialNB(client=client)
nb_model.fit(X, y)
assert nb_model.feature_count_ is not None
assert isinstance(nb_model.feature_count_, cupy.core.ndarray)