[Bug] fix fillna function on a single column fail

sdks/python/apache_beam/dataframe/frames_test.py

@@ -1025,6 +1025,17 @@ def test_series_fillna_series_as_value(self):
 
     self._run_test(lambda df, df2: df.A.fillna(df2.A), df, df2)
 
+  def test_dataframe_column_fillna_constant_as_value(self):
+    from apache_beam.dataframe import convert
+    from apache_beam.testing.util import assert_that
+    from apache_beam.testing.util import equal_to
+    with beam.Pipeline(None) as p:
+      pcoll = (
+          p | beam.Create([1.0, np.nan, -1.0]) | beam.Select(x=lambda x: x))
+      df = convert.to_dataframe(pcoll)
+      df_new = df['x'].fillna(0)
+      assert_that(convert.to_pcollection(df_new), equal_to([1.0, 0.0, -1.0]))
+
   @unittest.skipIf(PD_VERSION >= (2, 0), 'append removed in Pandas 2.0')
   def test_append_verify_integrity(self):
     df1 = pd.DataFrame({'A': range(10), 'B': range(10)}, index=range(10))

sdks/python/apache_beam/dataframe/transforms.py

@@ -395,7 +395,11 @@ def expr_to_stages(expr):
 
       if stage is None:
         # No stage available, compute this expression as part of a new stage.
-        stage = Stage(expr.args(), expr.requires_partition_by())
+        stage = Stage([
+            arg for arg in expr.args()
+            if not isinstance(arg, expressions.ConstantExpression)
+        ],
+                      expr.requires_partition_by())
         for arg in expr.args():
           # For each argument, declare that it is also available in
           # this new stage.