These are chat archives for astropy/astropy

27th
Mar 2016
Adam Ginsburg
@keflavich
Mar 27 2016 16:59
hi cdeil
Adam Ginsburg
@keflavich
Mar 27 2016 17:06
@astrofrog think we can merge this this morning? astrofrog/reproject#96 & possibly review & merge astropy/regions#10
Adrian Price-Whelan
@adrn
Mar 27 2016 17:18
RE astropy modeling discussion at the LSST/Astropy summit: The word "model" is so so so overloaded -- I try not to use it in papers unless it has a word of context (e.g., statistical model, gravitational potential model, etc.). I wonder if we can somehow get rid of this word
Thomas Robitaille
@astrofrog
Mar 27 2016 17:21
@adrn - do you have a suggested terminology?
@keflavich - there's a sphinx failure in astrofrog/reproject#96
We also need to add a unit test for the arbitrary reprojection
Adrian Price-Whelan
@adrn
Mar 27 2016 17:24
@astrofrog I'm not sure there is a good alternative, so in practice I've either had to default to adding qualifiers to the word "model" or invent new terminology -- both are probably bad?
Thomas Robitaille
@astrofrog
Mar 27 2016 17:24
@adrn - ok, so like always saying 'parametric models'?
or maybe parametric functions?
Adrian Price-Whelan
@adrn
Mar 27 2016 17:24
right
Adam Ginsburg
@keflavich
Mar 27 2016 17:25
always saying 'parametric models' soudns good; probably should rename the package to that (add that suggestion to the APE?)
Thomas Robitaille
@astrofrog
Mar 27 2016 17:25
@keflavich - can you rebase astropy/regions#10 to get rid of the merge commit?
Adam Ginsburg
@keflavich
Mar 27 2016 17:25
yes
Thomas Robitaille
@astrofrog
Mar 27 2016 17:26
@keflavich - I agree about adding something about terminology to the APE
Adam Ginsburg
@keflavich
Mar 27 2016 17:26
actually, @astrofrog, you're sure you want to remove the merge? I think that is actually a useful bit of history
and I am likely to break things in a rebase
it was not a clean merge
Adrian Price-Whelan
@adrn
Mar 27 2016 17:26
does the APE already exist as a document?
Thomas Robitaille
@astrofrog
Mar 27 2016 17:26
ok np, don't rebase
Adam Ginsburg
@keflavich
Mar 27 2016 17:26
cdeil said he was working on it
Adrian Price-Whelan
@adrn
Mar 27 2016 17:26
@cdeil
Adam Ginsburg
@keflavich
Mar 27 2016 17:27
…but I don't see it
Thomas Robitaille
@astrofrog
Mar 27 2016 17:27
@cdeil - once you have a first draft of APE, can you already open a WIP PR so we can add comments like this on Github?
Adrian Price-Whelan
@adrn
Mar 27 2016 17:27
+1
Thomas Robitaille
@astrofrog
Mar 27 2016 17:34
@keflavich - I reviewed astropy/regions#10, couple of small comments
Adam Ginsburg
@keflavich
Mar 27 2016 17:36
thx
Adam Ginsburg
@keflavich
Mar 27 2016 18:11
@astrofrog the warnings seem to exist with or without the added text in the docstring. Not sure how to debug this yet
Thomas Robitaille
@astrofrog
Mar 27 2016 18:11
ok will look
maybe it's a change in sphinx etc
Adam Ginsburg
@keflavich
Mar 27 2016 18:16
must be something about my local config
I have 79 warnings, but that's down from 80!
Brigitta Sipocz
@bsipocz
Mar 27 2016 18:21
@keflavich you mean warnings in astropy/regions#10?
Thomas Robitaille
@astrofrog
Mar 27 2016 18:22
Yes
Ah no sorry
we were talking about astrofrog/reproject#96
Adam Ginsburg
@keflavich
Mar 27 2016 18:24
@astrofrog I fixed the sphinx warning (it's green); turns out the other 79 are local and I don't understand them (maybe intersphinx failure?)
Brigitta Sipocz
@bsipocz
Mar 27 2016 18:25
if it's intersphinx related it should come up at the top where the intersphinx inventories are collected (sometimes there is an error there, but it's easy to overlook them as all the warnings are generated below)
Adam Ginsburg
@keflavich
Mar 27 2016 18:26
I don't see any such notes
or errors
Brigitta Sipocz
@bsipocz
Mar 27 2016 18:28
:+1:
Thomas Robitaille
@astrofrog
Mar 27 2016 19:20
@keflavich - are you planning on fixing the failures in astropy/regions#10
Brigitta Sipocz
@bsipocz
Mar 27 2016 19:21
@astrofrog - Ignore appveyor for that, I'm trying to fix it.
Adam Ginsburg
@keflavich
Mar 27 2016 19:21
yes
@bspiocz was your last comment (https://github.com/astropy/regions/pull/10#commitcomment-16858467) supposed to be on a specific line?
Brigitta Sipocz
@bsipocz
Mar 27 2016 19:23
Nope, it was a comment for the whole file. And doesn't really mean anything atm, but on the long term I think it would be nice to move this upstream into core (as there are a few similar functions in there already).
Adam Ginsburg
@keflavich
Mar 27 2016 19:24
yes
the plan is to have regions as a whole go into core
but first we want to do 1-2 versions
wait… the whole file, or whole PR? If file, which file? I think github did not link your comment to a file
Brigitta Sipocz
@bsipocz
Mar 27 2016 19:26
even better, but then it would be still great to have these wcs related functions in one place, and not scattered around in astropy.wcs and astropy.regions. But this is philosophical, and we shouldn't aim to solve it in this actual PR
the comment was for this commit/file keflavich/regions@6a0e2a5
Thomas Robitaille
@astrofrog
Mar 27 2016 19:55
@bsipocz @keflavich @adrn @cdeil - thoughts on astropy/regions#18 If so I can work on it while also adding the tests over all region types (since the registry would be useful for this)
Adam Ginsburg
@keflavich
Mar 27 2016 20:01
technical question for anyone (e.g., astrofrog): how do you set up ccache so that it stays first on the path when you activate an anaconda env?
Adam Ginsburg
@keflavich
Mar 27 2016 20:43
so, anything more on astropy/regions#10
Brigitta Sipocz
@bsipocz
Mar 27 2016 20:44
if @astrofrog happy I'm happy with it
Thomas Robitaille
@astrofrog
Mar 27 2016 23:13
@keflavich @cdeil @bsipocz we have our work cut out: https://github.com/astropy/regions/pull/20#issue-143845615 ;)
Adam Ginsburg
@keflavich
Mar 27 2016 23:16
niiice
talljimbo's gonna do that for us, right?
Thomas Robitaille
@astrofrog
Mar 27 2016 23:27
formatted table better
it all fits now