fix(post-processing): handle missing values in cumulative operator

superset/utils/pandas_postprocessing/cum.py

@@ -46,6 +46,7 @@ def cum(
     """
     columns = columns or {}
     df_cum = df.loc[:, columns.keys()]
+    df_cum = df_cum.fillna(0)
     operation = "cum" + operator
     if operation not in ALLOWLIST_CUMULATIVE_FUNCTIONS or not hasattr(
         df_cum, operation

tests/unit_tests/fixtures/dataframes.py

@@ -130,6 +130,11 @@
     data={"label": ["x", "y", "z", "q"], "y": [1.0, 2.0, 3.0, 4.0]},
 )
 
+timeseries_with_gap_df = DataFrame(
+    index=to_datetime(["2019-01-01", "2019-01-02", "2019-01-05", "2019-01-07"]),
+    data={"label": ["x", "y", "z", "q"], "y": [1.0, 2.0, None, 4.0]},
+)
+
 timeseries_df2 = DataFrame(
     index=to_datetime(["2019-01-01", "2019-01-02", "2019-01-05", "2019-01-07"]),
     data={

tests/unit_tests/pandas_postprocessing/test_cum.py

@@ -24,6 +24,7 @@
     multiple_metrics_df,
     single_metric_df,
     timeseries_df,
+    timeseries_with_gap_df,
 )
 from tests.unit_tests.pandas_postprocessing.utils import series_to_list
 
@@ -77,6 +78,19 @@ def test_cum():
         )
 
 
+def test_cum_with_gap():
+    # create new column (cumsum)
+    post_df = pp.cum(
+        df=timeseries_with_gap_df,
+        columns={"y": "y2"},
+        operator="sum",
+    )
+    assert post_df.columns.tolist() == ["label", "y", "y2"]
+    assert series_to_list(post_df["label"]) == ["x", "y", "z", "q"]
+    assert series_to_list(post_df["y"]) == [1.0, 2.0, None, 4.0]
+    assert series_to_list(post_df["y2"]) == [1.0, 3.0, 3.0, 7.0]
+
+
 def test_cum_after_pivot_with_single_metric():
     pivot_df = pp.pivot(
         df=single_metric_df,