rusackas commented on code in PR #43636: URL: https://github.com/apache/superset/pull/43636#discussion_r3878266906
########## docs/docs/using-superset/exploring-data.mdx: ########## @@ -151,6 +151,16 @@ see some data! You should see months in the rows and Department and Travel Class in the columns. Publish this chart to your existing Tutorial Dashboard you created earlier. +:::note +Row and column totals/subtotals for the Pivot Table are computed by the database, at the same +granularity as the total itself, rather than by summing the already-aggregated cells shown in the +table. This means totals are correct for non-additive metrics too, such as ratios Review Comment: Fixed — reworded to distinguish the two paths: additive metrics (SUM/COUNT/MIN/MAX) reduce client-side from the full-detail query, non-additive ones hit the database at each total's granularity. ########## docs/docs/using-superset/exploring-data.mdx: ########## @@ -151,6 +151,16 @@ see some data! You should see months in the rows and Department and Travel Class in the columns. Publish this chart to your existing Tutorial Dashboard you created earlier. +:::note +Row and column totals/subtotals for the Pivot Table are computed by the database, at the same +granularity as the total itself, rather than by summing the already-aggregated cells shown in the Review Comment: Fixed, same wording change as the codeant thread above. ########## docs/docs/using-superset/exploring-data.mdx: ########## @@ -151,6 +151,16 @@ see some data! You should see months in the rows and Department and Travel Class in the columns. Publish this chart to your existing Tutorial Dashboard you created earlier. +:::note +Row and column totals/subtotals for the Pivot Table are computed by the database, at the same +granularity as the total itself, rather than by summing the already-aggregated cells shown in the +table. This means totals are correct for non-additive metrics too, such as ratios +(`SUM(a)/SUM(b)`), `COUNT_DISTINCT`, `AVG`, and percentiles, not just additive ones like `SUM` or +`COUNT`. Because of this, there's no separate "Aggregation function" control for totals: a total +always reflects the metric's own definition evaluated at that total's level. The Table chart's +**Show summary** row behaves the same way for percentage metrics. Review Comment: Fixed — removed the "behaves the same way" claim and called out that the Table chart's Summary aggregation control can override a metric's own aggregation for the summary row, which the Pivot Table has no equivalent of. -- 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]
