Tweaks to sparkles / yoshi based on cycle 24 work

[taldcroft]

Description

This PR has a few small updates which facilitate the feasibility studies of cycle 24 proposed targets.

• Add target name to the report output
• Fix roll dev logic so max_roll_dev is really the max roll deviation when supplied.
  • It turns out there was a mistake in the API and one method (used only internally) had a roll_dev parameter that was basically redundant.
  • I made this deprecated and added a FutureWarning if it is supplied.
• Fixed some doc strings mistakes
• Allowed supplying additional proseco keyword args to run_one_yoshi and convert_yoshi_to_proseco_params. For the cycle 24 studies this allows passing target_name directly.

Interface impacts

roll_dev arg to AcaReviewTable.get_roll_intervals() method is now deprecated and will raise a FutureWarning if supplied.

Testing

Unit tests

• Mac

Independent check of unit tests by Jean

• Linux:

Functional tests

Works with https://gist.github.com/taldcroft/92d6d92ed80316a9a66dffbf9379eb52.

[jeanconn]

Should it also just get the target name if available in get_yoshi_params_from_ocat?

[taldcroft]

Should it also just get the target name if available in get_yoshi_params_from_ocat?

Yes, done.

[taldcroft]

@jeanconn - I fixed the issues you noted.

[jeanconn]

Thanks. I also ran this on the somewhat stressing case of obsid 17892 which has no target name. Did not cause any errors or issues with output. I note that in general that convert_yoshi_to_proseco_params doesn't behave well for old observations (< 2012) because chandra_aca.drift complains that the dynamic stuff doesn't work for historical observations that old. But that's not an issue for this code and not an issue for the way we are likely to use sparkles.