search

gammapy / gamma-cat

An open data collection and source catalog for gamma-ray astronomy
https://gamma-cat.readthedocs.io/
BSD 3-Clause "New" or "Revised" License
15 stars 17 forks source link

Pre- or posttrial significance? #106

Closed GernotMaier closed 7 years ago

GernotMaier commented 7 years ago

Which one do we usual quote in gamma-cat?

Noted this, as the following one uses the pre-trial significance:

https://github.com/gammapy/gamma-cat/blob/master/input/data/2013/2013ApJ...764...38A/tev-000001.yaml

cdeil commented 7 years ago

I always put pre-trial significance.

(Post-trial significance is a scam.)

I'll keep this issue open until this is mentioned in the gamma-cat docs.

GernotMaier commented 7 years ago

OK - I actually think that pre-trial significance is a useless numbers, without knowing the number of trials.

cdeil commented 7 years ago

The number of trials is only known if everything is fixed before observing (target, exposure time, cuts, background estimation, energy range, assumed spectrum, ...). I'm not aware of a single TeV measurement where that is the case.

Probably doesn't make sense to argue that one over Github ...

Should we add a significance_post_trial key (or something shorter if you have an idea) in the schema so that we have a place to put this number if it's given in the paper?

GernotMaier commented 7 years ago

significance_post_trial is good - better than having an abbreviation which is not clear.

cdeil commented 7 years ago

I've improved the schema files quite a bit, adding descriptions to most fields. And I've added significance_post_trial in 0aeac1e (not used by any dataset at the moment).