sadpandajoe commented on code in PR #44657:
URL: https://github.com/apache/superset/pull/44657#discussion_r4257458964
##########
tests/unit_tests/charts/test_client_processing.py:
##########
@@ -2144,6 +2172,190 @@ def
test_pivot_table_v2_actual_values_uses_database_rollups():
assert pivoted.loc[("US",), ("AVG(num)", "boy")] == 10
[email protected](
+ "aggregate_function,expected",
+ [
+ ("Sample Variance", 50.0),
+ ("Sample Standard Deviation", 50.0**0.5),
+ ],
+)
+def test_pivot_table_v2_sample_statistics_aggregations(
+ aggregate_function: str, expected: float
+):
+ """Sample variance/standard deviation must compute for two values at one
+ pivot address (10 and 20 give variance 50 and standard deviation ~7.071),
+ not raise on a misspelled ``pd.series`` or a non-callable tuple.
+ """
+ df = pd.DataFrame(
+ {
+ "nation": ["US", "US"],
+ "gender": ["boy", "boy"],
+ "SUM(num)": [10, 20],
+ }
+ )
+ form_data = {
+ "groupbyRows": ["nation"],
+ "groupbyColumns": ["gender"],
+ "metrics": ["SUM(num)"],
+ "aggregateFunction": aggregate_function,
+ }
+
+ pivoted = pivot_table_v2(df, form_data, apply_number_format=False)
+
+ assert pivoted.loc[("US",), ("SUM(num)", "boy")] == pytest.approx(expected)
+
+
[email protected](
+ "aggregate_function",
+ ["Sample Variance", "Sample Standard Deviation"],
+)
+def test_pivot_table_v2_sample_statistics_with_summaries(aggregate_function:
str):
+ """Row and column summaries invoke the reducer with ``axis``; the sample
+ statistics must handle the DataFrame/axis form, not just a Series.
+ """
+ df = pd.DataFrame(
+ {
+ "nation": ["US", "US", "FR", "FR"],
+ "gender": ["boy", "girl", "boy", "girl"],
+ "SUM(num)": [10, 20, 30, 50],
+ }
+ )
+ form_data = {
+ "groupbyRows": ["nation"],
+ "groupbyColumns": ["gender"],
+ "metrics": ["SUM(num)"],
+ "aggregateFunction": aggregate_function,
+ "rowTotals": True,
+ "colTotals": True,
+ }
+
+ pivoted = pivot_table_v2(df, form_data, apply_number_format=False)
+
+ assert not pivoted.empty
Review Comment:
This test still only asserts the pivot is non-empty, so a DataFrame/`axis`
reducer that returned 0 for every summary would pass. The #45215 gap only
collapses summaries when each (nation, gender) cell holds a single observation;
with two observations per cell, as suggested earlier in this thread, the
summaries are non-zero today. Could this use that fixture and assert the
concrete row and column variances (and square roots for the standard deviation)
now, rather than waiting on that issue?
--
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]
