fix

apache · Jun 7, 2024 · 63ef2d8 · 63ef2d8
1 parent 835a173
commit 63ef2d8
Show file tree

Hide file tree

Showing 5 changed files with 47 additions and 23 deletions.
diff --git a/python/pyspark/pandas/plot/core.py b/python/pyspark/pandas/plot/core.py
@@ -572,10 +572,14 @@ def _get_plot_backend(backend=None):
         return module
 
     def __call__(self, kind="line", backend=None, **kwargs):
+        kind = {"density": "kde"}.get(kind, kind)
+
+        if is_remote() and kind in ["hist", "kde"]:
+            return unsupported_function(class_name="pd.DataFrame", method_name=kind)()
+
         plot_backend = PandasOnSparkPlotAccessor._get_plot_backend(backend)
         plot_data = self.data
 
-        kind = {"density": "kde"}.get(kind, kind)
         if hasattr(plot_backend, "plot_pandas_on_spark"):
             # use if there's pandas-on-Spark specific method.
             return plot_backend.plot_pandas_on_spark(plot_data, kind=kind, **kwargs)

diff --git a/python/pyspark/pandas/tests/connect/plot/test_parity_frame_plot_matplotlib.py b/python/pyspark/pandas/tests/connect/plot/test_parity_frame_plot_matplotlib.py
@@ -18,19 +18,22 @@
 
 from pyspark.pandas.tests.plot.test_frame_plot_matplotlib import DataFramePlotMatplotlibTestsMixin
 from pyspark.testing.connectutils import ReusedConnectTestCase
+from pyspark.pandas.exceptions import PandasNotImplementedError
 from pyspark.testing.pandasutils import PandasOnSparkTestUtils, TestUtils
 
 
 class DataFramePlotMatplotlibParityTests(
     DataFramePlotMatplotlibTestsMixin, PandasOnSparkTestUtils, TestUtils, ReusedConnectTestCase
 ):
-    @unittest.skip("Test depends on Spark ML which is not supported from Spark Connect.")
     def test_hist_plot(self):
-        super().test_hist_plot()
+        # "Test depends on Spark ML which is not supported from Spark Connect."
+        with self.assertRaises(PandasNotImplementedError) as pe:
+            super().test_hist_plot()
 
-    @unittest.skip("Test depends on Spark ML which is not supported from Spark Connect.")
     def test_kde_plot(self):
-        super().test_kde_plot()
+        # "Test depends on Spark ML which is not supported from Spark Connect."
+        with self.assertRaises(PandasNotImplementedError) as pe:
+            super().test_kde_plot()
 
 
 if __name__ == "__main__":

diff --git a/python/pyspark/pandas/tests/connect/plot/test_parity_frame_plot_plotly.py b/python/pyspark/pandas/tests/connect/plot/test_parity_frame_plot_plotly.py
@@ -18,23 +18,27 @@
 
 from pyspark.pandas.tests.plot.test_frame_plot_plotly import DataFramePlotPlotlyTestsMixin
 from pyspark.testing.connectutils import ReusedConnectTestCase
+from pyspark.pandas.exceptions import PandasNotImplementedError
 from pyspark.testing.pandasutils import PandasOnSparkTestUtils, TestUtils
 
 
 class DataFramePlotPlotlyParityTests(
     DataFramePlotPlotlyTestsMixin, PandasOnSparkTestUtils, TestUtils, ReusedConnectTestCase
 ):
-    @unittest.skip("Test depends on Spark ML which is not supported from Spark Connect.")
     def test_hist_layout_kwargs(self):
-        super().test_hist_layout_kwargs()
+        # "Test depends on Spark ML which is not supported from Spark Connect."
+        with self.assertRaises(PandasNotImplementedError) as pe:
+            super().test_hist_layout_kwargs()
 
-    @unittest.skip("Test depends on Spark ML which is not supported from Spark Connect.")
     def test_hist_plot(self):
-        super().test_hist_plot()
+        # "Test depends on Spark ML which is not supported from Spark Connect."
+        with self.assertRaises(PandasNotImplementedError) as pe:
+            super().test_hist_plot()
 
-    @unittest.skip("Test depends on Spark ML which is not supported from Spark Connect.")
     def test_kde_plot(self):
-        super().test_kde_plot()
+        # "Test depends on Spark ML which is not supported from Spark Connect."
+        with self.assertRaises(PandasNotImplementedError) as pe:
+            super().test_kde_plot()
 
 
 if __name__ == "__main__":

diff --git a/python/pyspark/pandas/tests/connect/plot/test_parity_series_plot_matplotlib.py b/python/pyspark/pandas/tests/connect/plot/test_parity_series_plot_matplotlib.py
@@ -18,27 +18,37 @@
 
 from pyspark.pandas.tests.plot.test_series_plot_matplotlib import SeriesPlotMatplotlibTestsMixin
 from pyspark.testing.connectutils import ReusedConnectTestCase
+from pyspark.pandas.exceptions import PandasNotImplementedError
 from pyspark.testing.pandasutils import PandasOnSparkTestUtils, TestUtils
 
 
 class SeriesPlotMatplotlibParityTests(
     SeriesPlotMatplotlibTestsMixin, PandasOnSparkTestUtils, TestUtils, ReusedConnectTestCase
 ):
-    @unittest.skip("Test depends on Spark ML which is not supported from Spark Connect.")
+    def test_empty_hist(self):
+        # "Test depends on Spark ML which is not supported from Spark Connect."
+        with self.assertRaises(PandasNotImplementedError) as pe:
+            super().test_empty_hist()
+
     def test_hist(self):
-        super().test_hist()
+        # "Test depends on Spark ML which is not supported from Spark Connect."
+        with self.assertRaises(PandasNotImplementedError) as pe:
+            super().test_hist()
 
-    @unittest.skip("Test depends on Spark ML which is not supported from Spark Connect.")
     def test_hist_plot(self):
-        super().test_hist_plot()
+        # "Test depends on Spark ML which is not supported from Spark Connect."
+        with self.assertRaises(PandasNotImplementedError) as pe:
+            super().test_hist_plot()
 
-    @unittest.skip("Test depends on Spark ML which is not supported from Spark Connect.")
     def test_kde_plot(self):
-        super().test_kde_plot()
+        # "Test depends on Spark ML which is not supported from Spark Connect."
+        with self.assertRaises(PandasNotImplementedError) as pe:
+            super().test_kde_plot()
 
-    @unittest.skip("Test depends on Spark ML which is not supported from Spark Connect.")
     def test_single_value_hist(self):
-        super().test_single_value_hist()
+        # "Test depends on Spark ML which is not supported from Spark Connect."
+        with self.assertRaises(PandasNotImplementedError) as pe:
+            super().test_single_value_hist()
 
 
 if __name__ == "__main__":

diff --git a/python/pyspark/pandas/tests/connect/plot/test_parity_series_plot_plotly.py b/python/pyspark/pandas/tests/connect/plot/test_parity_series_plot_plotly.py
@@ -18,19 +18,22 @@
 
 from pyspark.pandas.tests.plot.test_series_plot_plotly import SeriesPlotPlotlyTestsMixin
 from pyspark.testing.connectutils import ReusedConnectTestCase
+from pyspark.pandas.exceptions import PandasNotImplementedError
 from pyspark.testing.pandasutils import PandasOnSparkTestUtils, TestUtils
 
 
 class SeriesPlotPlotlyParityTests(
     SeriesPlotPlotlyTestsMixin, PandasOnSparkTestUtils, TestUtils, ReusedConnectTestCase
 ):
-    @unittest.skip("Test depends on Spark ML which is not supported from Spark Connect.")
     def test_hist_plot(self):
-        super().test_hist_plot()
+        # "Test depends on Spark ML which is not supported from Spark Connect."
+        with self.assertRaises(PandasNotImplementedError) as pe:
+            super().test_hist_plot()
 
-    @unittest.skip("Test depends on Spark ML which is not supported from Spark Connect.")
     def test_kde_plot(self):
-        super().test_kde_plot()
+        # "Test depends on Spark ML which is not supported from Spark Connect."
+        with self.assertRaises(PandasNotImplementedError) as pe:
+            super().test_kde_plot()
 
 
 if __name__ == "__main__":