Use limits from the JSON file, refactor prediction plots

[jzuhone]

Description

This PR includes two refactors for acis_thermal_check and a bug fix.

1. Use limits from the JSON file

In sot/chandra_models#75 limit information was added to the model specification JSON files. In sot/xija#112 the ability to select a default color based on a limit was added. This PR uses both of these pieces of functionality in acis_thermal_check. This means that the limit information no longer needs to reside in a separate place in the limits.yml file. The limits also had to be added to the test JSON files for the regression testing suite.

2. Refactor prediction plotting

This PR also includes a refactor of prediction plotting. For making the prediction plots, we had been using two legacy functions, plot_one and plot_two, the first of which plotted one quantity (e.g. the temperature) on the left y-axis only, and the second of which added another quantity (e.g. the pitch) on the right y-axis. Much of the code between these two functions was duplicated. Also, these plots were handled internally and passed to the code which makes the web page through dictionaries, which creates quite a bit of extra code from the use of indexing the dicts. I have merged the functionalities of plot_one and plot_two into a new PredictPlot class, which also stores everything that was in a dict before as attributes, such as the Matplotlib Figure and Axes for the plot. This results in code which is much shorter and cleaner. I also added a method for this new class, add_limit_line, for adding a limit line to the plot to reduce code duplication.

This lays the groundwork for doing a similar thing later for the validation plots, but to avoid making this PR even larger I'll hold off on that for now.

3. This PR also fixes a bug which included obsids in the list of observations which are permitted to go to -109 C which are not part of the load under review.

Testing

• Passes unit tests on MacOS, linux, Windows (at least one required)
• Functional testing

[jzuhone]

@Gregg140 @jazan12 just a reminder that this needs looked at

[jazan12]

Included a few minor and superficial notes for you to check but looks good to me.

[jazan12]

Yes, that's right. Just want to make sure that if you're assessing the model prediction versus assessing which time counts as cold, different limits are being used.

…

On Mon, Jan 24, 2022 at 9:34 PM John ZuHone ***@***.***> wrote: ***@***.**** commented on this pull request. ------------------------------ In acis_thermal_check/tests/acisfp/acisfp_test_spec.json <#45 (comment)> : > @@ -441,6 +441,17 @@ 794 ] }, + "limits": { + "fptemp": { + "planning.data_quality.high.acisi": -112.0, + "planning.data_quality.high.aciss": -111.0, + "planning.data_quality.high.aciss_hot": -109.0, + "planning.data_quality.high.cold_ecs": -118.2, @jazan12 <https://github.com/jazan12> that's right, but in order to avoid hitting that limit, we plan to -118.2 C, correct? — Reply to this email directly, view it on GitHub <#45 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACN5TGKJPQSGS5NSTACFIF3UXYDYLANCNFSM5K4YVDSQ> . Triage notifications on the go with GitHub Mobile for iOS <https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675> or Android <https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub>. You are receiving this because you were mentioned.Message ID: ***@***.***>