wbo4958 commented on code in PR #49596:
URL: https://github.com/apache/spark/pull/49596#discussion_r1926309716
##########
python/pyspark/ml/tests/test_pipeline.py:
##########
@@ -58,6 +70,109 @@ def doTransform(pipeline):
# check that failure to set stages param will raise KeyError for
missing param
self.assertRaises(KeyError, lambda: doTransform(Pipeline()))
+ def test_classification_pipeline(self):
+ df = self.spark.createDataFrame([(1, 1, 0, 3), (0, 2, 0, 1)], ["y",
"a", "b", "c"])
+
+ assembler = VectorAssembler(outputCol="features")
+ assembler.setInputCols(["a", "b", "c"])
+
+ scaler = MaxAbsScaler(inputCol="features", outputCol="scaled_features")
+
+ lr = LogisticRegression(featuresCol="scaled_features", labelCol="y")
+ lr.setMaxIter(1)
+
+ pipeline = Pipeline(stages=[assembler, scaler, lr])
+ self.assertEqual(len(pipeline.getStages()), 3)
+
+ # Pipeline save & load
+ with tempfile.TemporaryDirectory(prefix="classification_pipeline") as
d:
+ pipeline.write().overwrite().save(d)
+ pipeline2 = Pipeline.load(d)
+ self.assertEqual(str(pipeline), str(pipeline2))
+ self.assertEqual(str(pipeline.getStages()),
str(pipeline2.getStages()))
+
+ model = pipeline.fit(df)
+ self.assertEqual(len(model.stages), 3)
+ self.assertIsInstance(model.stages[0], VectorAssembler)
+ self.assertIsInstance(model.stages[1], MaxAbsScalerModel)
+ self.assertIsInstance(model.stages[2], LogisticRegressionModel)
+
+ output = model.transform(df)
+ self.assertEqual(
+ output.columns,
+ [
+ "y",
+ "a",
+ "b",
+ "c",
+ "features",
+ "scaled_features",
+ "rawPrediction",
+ "probability",
+ "prediction",
+ ],
+ )
+ self.assertEqual(output.count(), 2)
+
+ # PipelineModel save & load
+ with
tempfile.TemporaryDirectory(prefix="classification_pipeline_model") as d:
+ model.write().overwrite().save(d)
+ model2 = PipelineModel.load(d)
+ self.assertEqual(str(model), str(model2))
+ self.assertEqual(str(model.stages), str(model2.stages))
+
+ def test_clustering_pipeline(self):
+ df = self.spark.createDataFrame([(1, 1, 0, 3), (0, 2, 0, 1)], ["y",
"a", "b", "c"])
+
+ assembler = VectorAssembler(outputCol="features")
+ assembler.setInputCols(["a", "b", "c"])
+
+ scaler = MinMaxScaler(inputCol="features", outputCol="scaled_features")
+
+ km = KMeans(k=2, maxIter=2)
+
+ pipeline = Pipeline(stages=[assembler, scaler, km])
+ self.assertEqual(len(pipeline.getStages()), 3)
+
+ # Pipeline save & load
+ with tempfile.TemporaryDirectory(prefix="clustering_pipeline") as d:
+ pipeline.write().overwrite().save(d)
Review Comment:
similar comments as above
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
