User has an experimental plan in the most general case where she has n samples to measure using m detectors/detector positions, and there are a series of measurements to be made on each sample, e.g., a temperature or a time series, or a position series or sthg.
What they would like to see at the beamline is any number (user selectable) from 0 to (n x m) plot window-clusters pop up showing intermediate or reduced data (similar to the current viz server but preferrably more granularly configurable). In each window-cluster they would like to see waterfall plots etc. of the series of curves (time/temperature) for a single sample on a single detector (i.e., the ith sample on the jth detector)
They would like the save server to save each of the (n x m) measurement series using the same rationale, so if we are using the file system there are (n x m) folders with the right data and metadata in them for each of the sample+detector series, all reduced to i(q), and if desired g(r), f(q), s(q) etc.. If we are using databroker/intake the same logic is captured somehow in the document structure (or at least in the catalog) allowing the user to fine-tune and redo analyses straightforwardly, maybe using xpdtools or sthg else.
There are, immediately, 4 UCs
UC1) 2 separate detectors at different positions exposed in an interleaved fashion, so the measurement logic looks like:
temperature 1:
sample 1:
detector1
detector2
sample 2:
detector1
detector2
temperature 2:
.....
UC2)
as UC1 but detector 2 is just detector 1 at a different position
UC3)
as UC1 but detector1 and detector 2 are offset sideways and triggered together
UC4)
as UC1 but the outer-loop is over the samples and the inner loop over the temperatures. The innermost loop is still over detectors.
multi-sample, multi-detector (or detector position) operations
Expected Behavior
User has an experimental plan in the most general case where she has n samples to measure using m detectors/detector positions, and there are a series of measurements to be made on each sample, e.g., a temperature or a time series, or a position series or sthg.
What they would like to see at the beamline is any number (user selectable) from 0 to (n x m) plot window-clusters pop up showing intermediate or reduced data (similar to the current viz server but preferrably more granularly configurable). In each window-cluster they would like to see waterfall plots etc. of the series of curves (time/temperature) for a single sample on a single detector (i.e., the ith sample on the jth detector)
They would like the save server to save each of the (n x m) measurement series using the same rationale, so if we are using the file system there are (n x m) folders with the right data and metadata in them for each of the sample+detector series, all reduced to i(q), and if desired g(r), f(q), s(q) etc.. If we are using databroker/intake the same logic is captured somehow in the document structure (or at least in the catalog) allowing the user to fine-tune and redo analyses straightforwardly, maybe using xpdtools or sthg else.
There are, immediately, 4 UCs UC1) 2 separate detectors at different positions exposed in an interleaved fashion, so the measurement logic looks like:
UC2) as UC1 but detector 2 is just detector 1 at a different position
UC3) as UC1 but detector1 and detector 2 are offset sideways and triggered together
UC4) as UC1 but the outer-loop is over the samples and the inner loop over the temperatures. The innermost loop is still over detectors.
Possible Solution
Steps to Reproduce (for bugs)
1. 2. 3.
Context
Priority
Your Environment