yousoph commented on PR #42145: URL: https://github.com/apache/superset/pull/42145#issuecomment-5039963227
Thanks for the review @rusackas! Fair point on the cadence. Some notes on the blast radius as shipped: the `P1D` fallback only engages when no grain is resolvable anywhere (panel control, x-axis column grain, or dashboard-applied extras), and the frequency only affects the appended *future* points from `make_future_dataframe` — the historical fit stays anchored to the actual data timestamps. So the worst case today is the forecast horizon extending at daily spacing over non-daily data (vs. the previous behavior for this state, which was the unhandled 500). Agreed it's worth a follow-up. The natural improvement is to infer the frequency from the data when the grain is unset — `pd.infer_freq` over the temporal index, falling back to the median timestamp delta, then `P1D` as a last resort — which would keep the forecast cadence aligned with non-daily/irregular series and also improve raw API callers that omit `time_grain`. Filed internally as SC-114844 to track it. -- 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]
