Filter scope on dashboard causes datetime filter to snap to first available data point instead of selected start

[seohyunjun]

When applying a datetime filter with filter scoping in a Superset dashboard, the affected chart does not respect the original start date set in the filter. Instead, the chart displays data starting from the first available data point in the dataset, which may be later than the selected filter start time.

This behavior leads to confusion when users expect to see a consistent time range across all charts based on the filter values.

Expected behavior:
Charts should display data starting from the exact datetime value specified in the filter, even if the data starts later (e.g., show empty periods before the first data point).

superset_record2.mov

Environment:
• Apache Superset version: 4.1.2
• Browser: Chrome

[dosubot[bot]]

Hi @seohyunjun! I'm Dosu and I’m helping the superset team.

I couldn't find any existing issues or discussions about datetime filters snapping to the first available data point in dashboards. Based on the code, Superset's datetime filter logic (both frontend and backend) is designed to respect the exact time range you specify in the filter—meaning it should send your selected start and end datetimes to the backend and use those for queries, regardless of whether your dataset has data for the entire range. There’s no built-in logic in the filter components or API to automatically snap the filter to the first available data point in your dataset; any such behavior would likely be coming from the chart’s query or visualization logic, not the filter itself (source, source).

If you want charts to show empty periods before the first data point (instead of starting at the first available record), you’ll need to ensure your chart’s query or visualization logic can render empty intervals for missing data. Superset’s filters and scoping system will pass the full time range, but the chart itself may need to be configured (or the dataset adjusted) to fill in those gaps.

If you’re seeing different behavior, it could be due to the chart type, dataset structure, or how missing time intervals are handled in your queries. There’s no setting in the filter or filter scoping UI to change this behavior. If you need more specific guidance, sharing the chart type and dataset structure could help narrow down the cause.

To reply, just mention @dosu.

---

^{How did I do? Good | Irrelevant | Incorrect | Verbose | Hallucination | Report 🐛 | Other}