msyavuz commented on PR #43718:
URL: https://github.com/apache/superset/pull/43718#issuecomment-5494004958

   Thanks — the metrics-on-rows case is fixed in 148ae88.
   
   The swap now applies unconditionally. The `columns and not rows` branch 
transposes once more on its own, but the totals are inserted *after* that 
branch and *before* the final transpose, so they were flipped there too. With 
`groupbyColumns: []` and `rowTotals`, the export is now `(2, 1)` matching the 
chart, rather than `(4, 1)` with phantom Subtotal/Total rows.
   
   Doing it unconditionally exposed an asymmetry underneath, worth flagging 
since it changes output beyond this layout. A metric column can hold 
non-numeric values — a literal `"NULL"` — and only the row totals coerced them. 
Reducing across columns instead concatenated strings, so a total over two empty 
cells read `"nannan"` and one over `"NULL"` read `"NULLnan"`. Both axes now 
coerce the way the sibling block already did, and those totals come out as 
numbers.
   
   The golden assertions in the three `..._null_...` tests are regenerated. 
Every value that moved is either the flipped axis or one of those concatenated 
strings — worth a glance to confirm you read them the same way.
   
   313 tests pass across `tests/unit_tests/charts/`, `utils/excel_tests.py` and 
`pandas_postprocessing/test_pivot.py`.


-- 
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]

Reply via email to