search

lightsources / BES-XPCS-Pilot

Work with BES Data Science Team to develop Python code to write NeXus data file with proposed NeXus definitions.
0 stars 0 forks source link

take results from APS 8-ID-I and write NeXus file #2

Closed prjemian closed 3 years ago

prjemian commented 4 years ago

@JulReinhardt mentioned in Slack today:

... in that example file from APS, there is an entry

entry/XPCS/data/masks/mask

which actually contains a 2D mask with the q-rings that are used for calculation in the APS matlab software. In xi-cam we would not need these for calculation in the future, but the example file lacks content in the

entry/XPCS/data/rois

which should IMO represent some ROI selected by the user.

ambarb commented 4 years ago

@prjemian @ronpandolfi I remember a discussion of a 3D array to represent ROIs (pixels, pixels, slices) or (slices, pixels, pixels). XPCS would just the require maybe just (pixels, pixels, 1) or (1, pixels, pixels). There was some concern for beamlines in that they just wanted 2D arrays when Suresh, Eric, Andrei, Andi, and QZ were present. You presented a good generic argument why others would want options for slices (energy sensitive stuff, etc). Do we want to revisit this?

For the coarse roi for dynamic Q (in terms of SAXS, the beamline files name this: "roi_mask" or "dqmap" (CHX vs 8ID). it is a 2D array.

I am looking for a PR or something that has this info. If anyone files the discussion in this comment, can you paste it below. attn @JulReinhardt

prjemian commented 3 years ago

The nx_loader_aps.py will do this. Closing this stale issue.