modification to p_astro, included mass gap

• p_astro reads mean values from emfollow/data to compute values

• mass gap category included

• unittests changed/added

@shaon.ghosh @shasvath.kapadia

added 2 commits

• 03e79a81 - 1 commit from branch emfollow:master
• f706453a - modification to p_astro, included mass gap

Compare with previous version

added Classification label

changed the description

mentioned in merge request !355 (merged)

Here are a few high-level comments.

1. The pastro_url configuration variable and the code changes related to it seem like they might be orthogonal to the addition of the mass gap category. Can you factor those changes into two separate merge requests?
2. These changes will require modifications to GraceDb (server and client) to support the added VOEvent field before they can be merged. Have you opened issues with @tanner.prestegard?
3. Related to comment 2, I suggest MassGap instead of MG.
4. Also related to comment 2, I suggest renaming Terr to Terrestrial for consistency with the VOEvent keywords before we merge this.

added 6 commits

• f706453a...14db1e32 - 5 commits from branch emfollow:master
• 20122c85 - modification to p_astro, included mass gap

Compare with previous version

If there are changes needed to gracedb-client, I need to know today specifically what they are because I am submitting the final release for ER14 tomorrow. To be able to do any sort of testing, I will also need server-side changes, too.

@tanner.prestegard, please add an optional MassGap parameter to the classification group.

@leo-singer, if you want to provide information about the units, data type, or description, please do. Otherwise, I will use no units, float, and no description.

@tanner.prestegard are these units for the MassGap category? In which case, I would add that these are probabilities and do not have units?

@deep.chatterjee in that case I take it the units are stat.probability in accordance with the other classification parameters (?)

Or maybe I should say the ucd field, which I have assumed is some type of unit specifier.

@tanner.prestegard if the units for the other classification parameters are stat.probability it should be the same for MassGap

BTW also for the properties the units are also the same as classification. @shaon.ghosh would you agree?

Yes, they are all probabilities

There are subtle differences but as far as units are concerned they are all probabilities.

thanks @shaon.ghosh . @tanner.prestegard let me know if there was anything else needed

It's up to you if you want to provide a description for the field or not. Here is an example VOEvent on gracedb-playground if you want to see what the description is like for other fields: https://gracedb-playground.ligo.org/api/superevents/MS190220t/files/MS190220t-2-Initial.xml,0