X-Axis Limit not respected using X-Axis Type Timescale

[aalvarell]

Current Behavior

I have two charts that use the same data (1 week of data) but one is intended to show only 1 day and the other 1 week.
I set the limits in the "X-Axis Limit" field:

but the 1-day limit does not work, it shows 1 week of data (all the data available in the base data). Daily data on the left, weekly on the right:

Expected Behavior

The time limits should be respected, and the chart with 1 day limit should show just 1 day worth of data.

Steps To Reproduce

1. Get 1 week data
2. Create 2 time charts, 1 showing only one day, the other 1 week of data
3. Display the charts

Environment

• Dashboard version: 1.28.0
• Node-RED version: 4.1.0
• Node.js version: 20.19.4
• npm version: 10.8.2
• Platform/OS: Nodered on Docker (https://hub.docker.com/r/nodered/node-red)
• Browser: Chrome

Have you provided an initial effort estimate for this issue?

I can not provide an initial effort estimate

[colinl]

The time range limit only works when adding samples onto an existing chart. If you send it a batch in one go it will show them all. Try using a Split node to split the array into individual messages before feeding to the chart, though I not going to guarantee that will do it.

[colinl]

Or send it an extra message with an empty payload after sending the batch may do it.

[colinl]

In fact it seems you are correct, even when new data are added the old points are not removed from the chart unless the page is refreshed. This is true for 1.27.0 up to 1.29.0

[aalvarell]

The time range limit only works when adding samples onto an existing chart. If you send it a batch in one go it will show them all. Try using a Split node to split the array into individual messages before feeding to the chart, though I not going to guarantee that will do it.

Ok, thanks. The documentation says nothing about this behavior, so I thought the limit was applied in every case.
I will try another approach, as splitting into thousands of individual messages to apply the limit seems overkill in this case.