Is your feature request related to a problem? Please describe.
Some additional features have been requested for the Inferencing page.
- Liquid tag and results
- Option to clear the graph
- Configuration for maximum graph retention. At 1/s, 3600 is far too long. This should be in
livex.cfg and set according to how fast the data is inferencing.
- Feedback on flatfield selection. It's unclear if it succeeded when a selection is made.
- Request for a preview of the flatfield frame(s) being selected. Not sure how feasible this is from a programming standpoint, but if this cannot be done then the above issue of flatfield feedback is definitely required.
- Graphs should show between 0 and 1 always instead of between the range of values
Describe the solution you'd like
The additional features would be nice, otherwise it is important to make sure that the flatfield is obvious in its success or failure for user clarity.
Describe alternatives you've considered
For the graph retention, this could be done programmatically based on the inferencing reporting its frames per second, so that users can specify a 'time window' to look back on in seconds instead of a number of readings to show.
Additional context
The inferencing will be subject to future change anyway. But as we are a way off from plotting squares on images then it's probably best to improve what exists as this will likely be used at the next two beamtimes in a similar form.
Is your feature request related to a problem? Please describe.
Some additional features have been requested for the Inferencing page.
livex.cfgand set according to how fast the data is inferencing.Describe the solution you'd like
The additional features would be nice, otherwise it is important to make sure that the flatfield is obvious in its success or failure for user clarity.
Describe alternatives you've considered
For the graph retention, this could be done programmatically based on the inferencing reporting its frames per second, so that users can specify a 'time window' to look back on in seconds instead of a number of readings to show.
Additional context
The inferencing will be subject to future change anyway. But as we are a way off from plotting squares on images then it's probably best to improve what exists as this will likely be used at the next two beamtimes in a similar form.