eschutho opened a new pull request, #40936:
URL: https://github.com/apache/superset/pull/40936
## What
Silences a pandas `FutureWarning` that fires on every chart render using
time-shift joins:
```
FutureWarning: Series.__getitem__ treating keys as positions is deprecated.
In a future version, integer keys will always be treated as labels
(consistent with DataFrame behavior). To access a value by position,
use `ser.iloc[pos]`
value = row[column_index]
```
**Source:** `superset/models/helpers.py` —
`ExploreMixin.generate_join_column`
## Change
```diff
- value = row[column_index]
+ value = row.iloc[column_index]
```
`df.apply(..., axis=1)` produces a `pd.Series` per row whose index labels
are the DataFrame column names (always strings from SQL result sets). `row[0]`
was being resolved positionally, which pandas now warns against.
`.iloc[column_index]` is the explicit positional API and matches the original
intent exactly.
## No behavior change
`row[0]` and `row.iloc[0]` return the same value for any DataFrame whose
columns are string-labeled (the only case that can reach this code path). This
is a pure deprecation fix.
## Test plan
- Existing tests in `tests/unit_tests/common/test_time_shifts.py` cover
`generate_join_column` via all four time grains and pass on this branch.
- No new tests needed: the integer-label edge case cannot arise in
production (all SQL result set columns are string-named).
--
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]
