import os
sys.path.insert(0, sys.argv[1])
os.environ['PYSPARK_PYTHON'] = sys.executable
import unittest
from pysparkling.context import H2OContext
from pysparkling.conf import H2OConf
from pyspark.sql import SparkSession
import unit_test_utils
import generic_test_utils
class H2OConfTest(unittest.TestCase):
@classmethod
def setUpClass(cls):
cls._cloud_name = generic_test_utils.unique_cloud_name("h2o_conf_test")
cls._conf = unit_test_utils.get_default_spark_conf(cls._spark_options_from_params). \
set("spark.ext.h2o.cloud.name", cls._cloud_name)
cls._spark = SparkSession.builder.config(conf=cls._conf).getOrCreate()
cls._hc = H2OContext.getOrCreate(cls._spark, H2OConf(cls._spark).set_num_of_external_h2o_nodes(1))
# test passing h2o_conf to H2OContext
def test_h2o_conf(self):
self.assertEquals(self._hc.get_conf().cloud_name(), self._cloud_name,
"Configuration property cloud_name should match")
if __name__ == '__main__':
generic_test_utils.run_tests([H2OConfTest], file_name="py_unit_tests_conf_report")
df = self._spark.createDataFrame(data)
hf = hc.as_h2o_frame(df)
# Modify H2O frame - this should invalidate internal cache
hf['c3'] = 3
# Now try to convert modified H2O frame back to Spark data frame
dfe = hc.as_spark_frame(hf)
self.assertEquals(dfe.count(), len(data),
"Number of rows should match")
self.assertEquals(len(dfe.columns), 3,
"Number of columns should match")
self.assertEquals(
dfe.collect(),
[Row(c1=1, c2='first', c3=3),
Row(c1=2, c2='second', c3=3)])
def test_sparse_data_conversion(self):
data = [(float(x), SparseVector(50000, {x: float(x)}))
for x in range(1, 90)]
df = self._spark.sparkContext.parallelize(data).toDF()
t0 = time.time()
self._hc.as_h2o_frame(df)
t1 = time.time()
total = t1 - t0
assert total < 10 # The conversion should not take longer then 10 seconds
if __name__ == '__main__':
generic_test_utils.run_tests([FrameTransformationsTest],
file_name="py_unit_tests_conversions_report")
sys.path.insert(0, sys.argv[1])
os.environ['PYSPARK_PYTHON'] = sys.executable
os.environ['PYSPARK_DRIVER_PYTHON'] = sys.executable
from integ_test_utils import *
from generic_test_utils import run_tests
class YarnIntegTestSuite(unittest.TestCase):
@classmethod
def setUpClass(cls):
conf = get_default_spark_conf(cls._spark_options_from_params)
conf["spark.master"] = "local[*]"
conf["spark.submit.pyFiles"] = sys.argv[1]
# Configure YARN environment
conf["spark.yarn.max.executor.failures"] = "1" # In fail of executor, fail the test
conf["spark.executor.instances"] = "1"
cls._conf = conf
def test_xgboost_medium(self):
return_code = launch(self._conf, "examples/scripts/tests/xgboost_test_medium.py")
self.assertTrue(return_code == 0, "Process ended in a wrong way. It ended with return code " + str(return_code))
def test_chicago_crime(self):
return_code = launch(self._conf, "examples/scripts/ChicagoCrimeDemo.py")
self.assertTrue(return_code == 0, "Process ended in a wrong way. It ended with return code " + str(return_code))
if __name__ == '__main__':
run_tests([YarnIntegTestSuite], file_name="py_integ_yarn_tests_report")
class H2OMojoPredictionsTest(unittest.TestCase):
@classmethod
def setUpClass(cls):
cls._cloud_name = generic_test_utils.unique_cloud_name("h2o_mojo_predictions_test")
cls._spark = SparkSession.builder.config(conf = unit_test_utils.get_default_spark_conf()).getOrCreate()
# test predictions on H2O Mojo
def test_h2o_mojo_predictions(self):
# Try loading the Mojo and prediction on it without starting H2O Context
mojo = H2OMOJOModel.create_from_mojo("../ml/src/test/resources/binom_model_prostate.mojo")
prostate_frame = self._spark.read.csv("file://" + unit_test_utils.locate("smalldata/prostate/prostate.csv"), header=True)
mojo.predict(prostate_frame).repartition(1).collect()
def test_h2o_mojo_predictions_unseen_categoricals(self):
mojo = H2OMOJOModel.create_from_mojo("../ml/src/test/resources/deep_learning_airlines_categoricals.zip")
mojo.setConvertUnknownCategoricalLevelsToNa(True)
d =[{'sepal_len':5.1, 'sepal_wid':3.5, 'petal_len':1.4, 'petal_wid':0.2, 'class':'Missing_categorical'}]
df = self._spark.createDataFrame(d)
data = mojo.transform(df).collect()[0]
assert data["class"] == "Missing_categorical"
assert data["petal_len"] == 1.4
assert data["petal_wid"] == 0.2
assert data["sepal_len"] == 5.1
assert data["sepal_wid"] == 3.5
assert data["prediction_output"][0] == 5.240174068202646
if __name__ == '__main__':
generic_test_utils.run_tests([H2OMojoPredictionsTest], file_name="py_unit_tests_mojo_predictions_report")
@classmethod
def setUpClass(cls):
conf = get_default_spark_conf(cls._spark_options_from_params)
conf["spark.master"] = "local[*]"
conf["spark.submit.pyFiles"] = sys.argv[1]
# Configure YARN environment
conf[
"spark.yarn.max.executor.failures"] = "1" # In fail of executor, fail the test
conf["spark.executor.instances"] = "1"
cls._conf = conf
def test_xgboost_medium(self):
return_code = launch(self._conf,
"examples/scripts/tests/xgboost_test_medium.py")
self.assertTrue(
return_code == 0,
"Process ended in a wrong way. It ended with return code " +
str(return_code))
def test_chicago_crime(self):
return_code = launch(self._conf,
"examples/scripts/ChicagoCrimeDemo.py")
self.assertTrue(
return_code == 0,
"Process ended in a wrong way. It ended with return code " +
str(return_code))
if __name__ == '__main__':
run_tests([YarnIntegTestSuite], file_name="py_integ_yarn_tests_report")
def test_s3n_import(self):
fr = h2o.import_file("s3n://data.h2o.ai/h2o-open-tour/2016-nyc/weather.csv")
assert fr.ncol == 27
assert fr.nrow == 9768
def test_s3a_import(self):
fr = h2o.import_file("s3a://data.h2o.ai/h2o-open-tour/2016-nyc/weather.csv")
assert fr.ncol == 27
assert fr.nrow == 9768
def s3_import_export(self, scheme):
local_frame = h2o.import_file("/home/0xdiag/smalldata/logreg/prostate.csv")
timestamp = datetime.today().utcnow().strftime("%Y%m%d-%H%M%S")
unique_suffix = str(uuid.uuid4())
s3_path = scheme + "://test.0xdata.com/h2o-hadoop-tests/test-export/" + scheme + "/exported." + \
timestamp + "." + unique_suffix + ".csv.zip"
h2o.export_file(local_frame, s3_path)
s3_frame = h2o.import_file(s3_path)
assert_frame_equal(local_frame.as_data_frame(), s3_frame.as_data_frame())
def test_s3a_import_export(self):
self.s3_import_export("s3a")
@unittest.skip("skip")
def test_s3n_import_export(self):
self.s3_import_export("s3n")
if __name__ == '__main__':
generic_test_utils.run_tests([HadoopSmokeTestSuite], file_name="py_hadoop_smoke_tests_report")
pipeline.write().overwrite().save(
"file://" +
os.path.abspath("build/test_dai_pipeline_as_spark_pipeline"))
loaded_pipeline = Pipeline.load(
"file://" +
os.path.abspath("build/test_dai_pipeline_as_spark_pipeline"))
## Train the pipeline model
model = loaded_pipeline.fit(prostate_frame)
model.write().overwrite().save(
"file://" +
os.path.abspath("build/test_dai_pipeline_as_spark_pipeline_model"))
loaded_model = PipelineModel.load(
"file://" +
os.path.abspath("build/test_dai_pipeline_as_spark_pipeline_model"))
preds = loaded_model.transform(prostate_frame).repartition(1).select(
mojo.select_prediction_udf("AGE")).take(5)
assert preds[0][0] == 65.36320409515132
assert preds[1][0] == 64.96902128114817
assert preds[2][0] == 64.96721023747583
assert preds[3][0] == 65.78772654671035
assert preds[4][0] == 66.11327967814829
if __name__ == '__main__':
generic_test_utils.run_tests(
[H2OMojoPipelineTest], file_name="py_unit_tests_mojo_pipeline_report")
def test_s3a_import(self):
fr = h2o.import_file(
"s3a://data.h2o.ai/h2o-open-tour/2016-nyc/weather.csv")
assert fr.ncol == 27
assert fr.nrow == 9768
def s3_import_export(self, scheme):
local_frame = h2o.import_file(
"/home/0xdiag/smalldata/logreg/prostate.csv")
timestamp = datetime.today().utcnow().strftime("%Y%m%d-%H%M%S")
unique_suffix = str(uuid.uuid4())
s3_path = scheme + "://test.0xdata.com/h2o-hadoop-tests/test-export/" + scheme + "/exported." + \
timestamp + "." + unique_suffix + ".csv.zip"
h2o.export_file(local_frame, s3_path)
s3_frame = h2o.import_file(s3_path)
assert_frame_equal(local_frame.as_data_frame(),
s3_frame.as_data_frame())
def test_s3a_import_export(self):
self.s3_import_export("s3a")
@unittest.skip("skip")
def test_s3n_import_export(self):
self.s3_import_export("s3n")
if __name__ == '__main__':
generic_test_utils.run_tests([HadoopSmokeTestSuite],
file_name="py_hadoop_smoke_tests_report")
env = IntegTestEnv()
env.set_spark_master("local[*]")
env.conf("spark.ext.h2o.port.base", 63331)
return_code = launch(env,
"examples/pipelines/ham_or_spam_multi_algo.py",
"automl")
self.assertTrue(
return_code == 0,
"Process ended in a wrong way. It ended with return code " +
str(return_code))
def test_import_pysparkling_standalone_app(self):
env = IntegTestEnv()
env.set_spark_master("local[*]")
env.conf("spark.ext.h2o.port.base", 63331)
return_code = launch(
env,
"examples/scripts/tests/pysparkling_ml_import_overrides_spark_test.py"
)
self.assertTrue(
return_code == 0,
"Process ended in a wrong way. It ended with return code " +
str(return_code))
if __name__ == '__main__':
generic_test_utils.run_tests([LocalIntegTestSuite],
file_name="py_integ_local_tests_report")
env.set_spark_master("local[*]")
# Configure YARN environment
env.conf("spark.yarn.max.executor.failures", 1) # In fail of executor, fail the test
env.conf("spark.executor.instances", 1)
env.conf("spark.executor.memory", "2g")
env.conf("spark.ext.h2o.port.base", 63331)
env.conf("spark.driver.memory", "2g")
return_code = launch(env, "examples/scripts/tests/xgboost_test_medium.py")
self.assertTrue(return_code == 0, "Process ended in a wrong way. It ended with return code "+str(return_code))
def test_chicago_crime(self):
env = IntegTestEnv()
env.set_spark_master("local[*]")
# Configure YARN environment
env.conf("spark.yarn.max.executor.failures", 1) # In fail of executor, fail the test
env.conf("spark.executor.instances", 1)
env.conf("spark.executor.memory", "2g")
env.conf("spark.ext.h2o.port.base", 63331)
env.conf("spark.driver.memory", "2g")
return_code = launch(env, "examples/scripts/ChicagoCrimeDemo.py")
self.assertTrue(return_code == 0, "Process ended in a wrong way. It ended with return code "+str(return_code))
if __name__ == '__main__':
generic_test_utils.run_tests([YarnIntegTestSuite], file_name="py_integ_yarn_tests_report")
featuresCols=["sepal_len", "sepal_wid", "petal_len", "petal_wid"],
withDetailedPredictionCol=True)
model = algo.fit(self.dataset)
transformed = model.transform(self.dataset)
self.assertEquals(
transformed.select("detailed_prediction.cluster").head()[0], 0,
"Prediction should match")
self.assertEquals(
len(transformed.select("detailed_prediction.distances").head()[0]),
3, "Size of distances array should match")
def testUserPoints(self):
algo = H2OKMeans(
splitRatio=0.8,
seed=1,
k=3,
featuresCols=["sepal_len", "sepal_wid", "petal_len", "petal_wid"],
userPoints=[[4.9, 3.0, 1.4, 0.2], [5.6, 2.5, 3.9, 1.1],
[6.5, 3.0, 5.2, 2.0]])
model = algo.fit(self.dataset)
self.assertEquals(
model.transform(self.dataset).select("prediction").head()[0], 0,
"Prediction should match")
if __name__ == '__main__':
generic_test_utils.run_tests([H2OKMeansTestSuite],
file_name="py_unit_tests_kmeans_report")
#
# Licensed to the Apache Software Foundation (ASF) under one or more
# contributor license agreements. See the NOTICE file distributed with
# this work for additional information regarding copyright ownership.
# The ASF licenses this file to You under the Apache License, Version 2.0
# (the "License"); you may not use this file except in compliance with
# the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
#
"""
Integration tests for pySparkling for spark running in Standalone mode
"""
import generic_test_utils
from integ_test_utils import *
import unittest
class StandaloneIntegTestSuite(unittest.TestCase):
pass
if __name__ == '__main__':
generic_test_utils.run_tests([StandaloneIntegTestSuite], file_name="py_integ_standalone_tests_report")
return_code = launch(self._conf,
"examples/pipelines/ham_or_spam_multi_algo.py",
param="xgboost")
self.assertTrue(
return_code == 0,
"Process ended in a wrong way. It ended with return code " +
str(return_code))
def test_pipeline_automl(self):
return_code = launch(self._conf,
"examples/pipelines/ham_or_spam_multi_algo.py",
param="automl")
self.assertTrue(
return_code == 0,
"Process ended in a wrong way. It ended with return code " +
str(return_code))
def test_import_pysparkling_standalone_app(self):
return_code = launch(
self._conf,
"examples/scripts/tests/pysparkling_ml_import_overrides_spark_test.py"
)
self.assertTrue(
return_code == 0,
"Process ended in a wrong way. It ended with return code " +
str(return_code))
if __name__ == '__main__':
run_tests([LocalIntegTestSuite], file_name="py_integ_local_tests_report")
def testProducedMOJOModelAndLoadedMOJOModelReturnsSameResult(self):
targetEncoder = H2OTargetEncoder(labelCol="CAPSULE", inputCols=["RACE", "DPROS", "DCAPS"])
pipeline = Pipeline(stages=[targetEncoder])
producedModel = pipeline.fit(self._trainingDataset)
path = "file://" + os.path.abspath("build/testProducedMOJOModelAndLoadedMOJOModelReturnsSameResult")
producedModel.write().overwrite().save(path)
loadedModel = PipelineModel.load(path)
transformedByProducedModel = producedModel.transform(self._testingDataset)
transformedByLoadedModel = loadedModel.transform(self._testingDataset)
unit_test_utils.assert_data_frames_are_identical(transformedByProducedModel, transformedByLoadedModel)
def testTargetEncoderModelWithDisabledNoiseAndTargetEncoderMOJOModelTransformTheTrainingDatasetSameWay(self):
targetEncoder = H2OTargetEncoder()\
.setInputCols(["RACE", "DPROS", "DCAPS"])\
.setLabelCol("CAPSULE")\
.setHoldoutStrategy("None")\
.setNoise(0.0)
targetEncoderModel = targetEncoder.fit(self._trainingDataset)
transformedByModel = targetEncoderModel.transformTrainingDataset(self._trainingDataset)
transformedByMOJOModel = targetEncoderModel.transform(self._trainingDataset)
unit_test_utils.assert_data_frames_are_identical(transformedByModel, transformedByMOJOModel)
if __name__ == '__main__':
generic_test_utils.run_tests([H2OTargetEncoderTestSuite], file_name="py_unit_tests_target_encoder_report")
@classmethod
def setUpClass(cls):
conf = get_default_spark_conf(cls._spark_options_from_params)
conf["spark.master"] = "local[*]"
conf["spark.submit.pyFiles"] = sys.argv[1]
cls._conf = conf
def test_pipeline_gbm_mojo(self):
return_code = launch(self._conf, "examples/pipelines/ham_or_spam_multi_algo.py", param="gbm")
self.assertTrue(return_code == 0, "Process ended in a wrong way. It ended with return code " + str(return_code))
def test_pipeline_deep_learning(self):
return_code = launch(self._conf, "examples/pipelines/ham_or_spam_multi_algo.py", param="dl")
self.assertTrue(return_code == 0, "Process ended in a wrong way. It ended with return code " + str(return_code))
def test_pipeline_xgboost(self):
return_code = launch(self._conf, "examples/pipelines/ham_or_spam_multi_algo.py", param="xgboost")
self.assertTrue(return_code == 0, "Process ended in a wrong way. It ended with return code " + str(return_code))
def test_pipeline_automl(self):
return_code = launch(self._conf, "examples/pipelines/ham_or_spam_multi_algo.py", param="automl")
self.assertTrue(return_code == 0, "Process ended in a wrong way. It ended with return code " + str(return_code))
def test_import_pysparkling_standalone_app(self):
return_code = launch(self._conf, "examples/scripts/tests/pysparkling_ml_import_overrides_spark_test.py")
self.assertTrue(return_code == 0, "Process ended in a wrong way. It ended with return code " + str(return_code))
if __name__ == '__main__':
run_tests([LocalIntegTestSuite], file_name="py_integ_local_tests_report")
