Cumulative passing / failing positioners

[joesilber]

Bin the fit errors vs time for results from the fit_posparams run. Generate a single overview plot saying how many robots were working on each given day. Probably output some table file giving details on which positioners failed when, format TBD.

[sbailey]

This looks good, thanks, with a minor comment that the plot_posparams help string mentions "fit_params" which I think is "fit_posparams", right?

Please add a section to README.md or to a new file in doc/posparams.md (or similar) with examples of how to use this chain. I've lost the big picture context.

Working backwards from here:

• plot_posparams help string says takes the input csv file containing "merged" data from fit_params (fit_posparams?)
• fit_posparams takes the input Mnnnn.csv files, presumably like those in /global/cfs/cdirs/desi/engineering/fvc/desimeter/v0.2.0/positioners/?
• what generates those Mnnnn.csv files?
• I presume they trace to some analysis of the fvc-*.csv files in /global/cfs/cdirs/desi/engineering/fvc/desimeter/v0.2.0, generated by desi_fvc_proc, which takes a raw FVC image as input.

I'm sure the answers are in a combination of PRs and emails and slack messages, but let's get a basic end-to-end example documented here for the context of how the pieces fit together and what they do. That could happen before or after merging this PR, but let's not let it linger. i.e. if you don't document it before meging, please open an issue about adding those docs. If the answer to all of this is already in a single place like a DocDB entry, ok to just reference that instead in the README.

Thanks.

[joesilber]

Good point -- I added an issue capturing this need for README, with a link to your comment here. Suggest proceeding with merging this particular branch.