From petr at kubanek.net  Wed Jul  4 21:21:56 2018
From: petr at kubanek.net (Petr =?ISO-8859-1?Q?Kub=E1nek?=)
Date: Wed, 04 Jul 2018 18:21:56 -0700
Subject: [AstroPy] Parse DETSEC etc..
Message-ID: <250df15ff0ed43383d0ac4c9fe8b43dafddabc6f.camel@kubanek.net>

Hi Astropy,

I might be stupid, but I don't see any way how to parse DETSEC and
similar FITS header you can encounter in (multiframe) FITS files. I of
course know how to parse that with e.g. re and mapping result to float
array, but is there a better way? See for the code (yes, that will
become a function..unless I know about better way):

from astropy.io import fits
import re

ff=fits.open('file.fits')
r = re.compile('\[(\d+):(\d+),(\d+):(\d+)\]')
g = r.match(ff[1].header['DETSEC'])  # DETSEC is '[20:30,40:50]'
b=map(float, g.groups())
print(b)
[20,30,40,50]


Thanks

Petr Kub?nek
http://rts2.org

From erwin at mpe.mpg.de  Thu Jul  5 02:51:35 2018
From: erwin at mpe.mpg.de (Peter Erwin)
Date: Thu, 5 Jul 2018 08:51:35 +0200
Subject: [AstroPy] Parse DETSEC etc..
In-Reply-To: <250df15ff0ed43383d0ac4c9fe8b43dafddabc6f.camel@kubanek.net>
References: <250df15ff0ed43383d0ac4c9fe8b43dafddabc6f.camel@kubanek.net>
Message-ID: <3F38B7B1-268C-4274-B481-0369AB1F8E49@mpe.mpg.de>

If you know the names of the headers to parse and you know they will always
have the format ?[x1:x2,y1:y2]?, then a slightly shorter (and possibly more
idiomatic) way, which avoids regex and mapping, would be:

ff=fits.open('file.fits')
h = ff[1].header[?DETSEC?]
b = [int(x2) for x1 in h[1:-1].split(",") for x2 in a.split(":?)]

(Substitute ?float? for ?int? if you really want floating-point values; I?m
assuming that you might want to use the values from DETSEC to index
the data array, in which case they should be integers.)

But that may not be as flexible as what you want.


cheers,

Peter

> On Jul 5, 2018, at 3:21 AM, Petr Kub?nek <petr at kubanek.net> wrote:
> 
> Hi Astropy,
> 
> I might be stupid, but I don't see any way how to parse DETSEC and
> similar FITS header you can encounter in (multiframe) FITS files. I of
> course know how to parse that with e.g. re and mapping result to float
> array, but is there a better way? See for the code (yes, that will
> become a function..unless I know about better way):
> 
> from astropy.io import fits
> import re
> 
> ff=fits.open('file.fits')
> r = re.compile('\[(\d+):(\d+),(\d+):(\d+)\]')
> g = r.match(ff[1].header['DETSEC'])  # DETSEC is '[20:30,40:50]'
> b=map(float, g.groups())
> print(b)
> [20,30,40,50]
> 
> 
> Thanks
> 
> Petr Kub?nek
> http://rts2.org
> _______________________________________________
> AstroPy mailing list
> AstroPy at python.org
> https://mail.python.org/mailman/listinfo/astropy

=============================================================
Peter Erwin                   Max-Planck-Insitute for Extraterrestrial 
erwin at mpe.mpg.de              Physics, Giessenbachstrasse
tel. +49 (0)176 2481 7713     85748 Garching, Germany
fax  +49 (0)89 30000 3495     http://www.mpe.mpg.de/~erwin




From erwin at mpe.mpg.de  Thu Jul  5 03:41:12 2018
From: erwin at mpe.mpg.de (Peter Erwin)
Date: Thu, 5 Jul 2018 09:41:12 +0200
Subject: [AstroPy] Parse DETSEC etc..
In-Reply-To: <3F38B7B1-268C-4274-B481-0369AB1F8E49@mpe.mpg.de>
References: <250df15ff0ed43383d0ac4c9fe8b43dafddabc6f.camel@kubanek.net>
 <3F38B7B1-268C-4274-B481-0369AB1F8E49@mpe.mpg.de>
Message-ID: <49A7423A-DFB9-447A-B37E-5C6A0811A414@mpe.mpg.de>

And, of course, the list comprehension line should look something more like this:

b = [int(q) for p in h[1:-1].split(',') for q in p.split(':?)]

(that?ll teach me to change the intermediate variable names halfway through
writing the post?)

   ? Peter

> On Jul 5, 2018, at 8:51 AM, Peter Erwin <erwin at mpe.mpg.de> wrote:
> 
> If you know the names of the headers to parse and you know they will always
> have the format ?[x1:x2,y1:y2]?, then a slightly shorter (and possibly more
> idiomatic) way, which avoids regex and mapping, would be:
> 
> ff=fits.open('file.fits')
> h = ff[1].header[?DETSEC?]
> b = [int(x2) for x1 in h[1:-1].split(",") for x2 in a.split(":?)]
> 
> (Substitute ?float? for ?int? if you really want floating-point values; I?m
> assuming that you might want to use the values from DETSEC to index
> the data array, in which case they should be integers.)
> 
> But that may not be as flexible as what you want.
> 
> 
> cheers,
> 
> Peter
> 
>> On Jul 5, 2018, at 3:21 AM, Petr Kub?nek <petr at kubanek.net> wrote:
>> 
>> Hi Astropy,
>> 
>> I might be stupid, but I don't see any way how to parse DETSEC and
>> similar FITS header you can encounter in (multiframe) FITS files. I of
>> course know how to parse that with e.g. re and mapping result to float
>> array, but is there a better way? See for the code (yes, that will
>> become a function..unless I know about better way):
>> 
>> from astropy.io import fits
>> import re
>> 
>> ff=fits.open('file.fits')
>> r = re.compile('\[(\d+):(\d+),(\d+):(\d+)\]')
>> g = r.match(ff[1].header['DETSEC'])  # DETSEC is '[20:30,40:50]'
>> b=map(float, g.groups())
>> print(b)
>> [20,30,40,50]
>> 
>> 
>> Thanks
>> 
>> Petr Kub?nek
>> http://rts2.org
>> _______________________________________________
>> AstroPy mailing list
>> AstroPy at python.org
>> https://mail.python.org/mailman/listinfo/astropy
> 
> =============================================================
> Peter Erwin                   Max-Planck-Insitute for Extraterrestrial 
> erwin at mpe.mpg.de              Physics, Giessenbachstrasse
> tel. +49 (0)176 2481 7713     85748 Garching, Germany
> fax  +49 (0)89 30000 3495     http://www.mpe.mpg.de/~erwin
> 
> 
> 
> _______________________________________________
> AstroPy mailing list
> AstroPy at python.org
> https://mail.python.org/mailman/listinfo/astropy

=============================================================
Peter Erwin                   Max-Planck-Insitute for Extraterrestrial 
erwin at mpe.mpg.de              Physics, Giessenbachstrasse
tel. +49 (0)176 2481 7713     85748 Garching, Germany
fax  +49 (0)89 30000 3495     http://www.mpe.mpg.de/~erwin




From te.pickering at gmail.com  Thu Jul  5 11:14:28 2018
From: te.pickering at gmail.com (Timothy Pickering)
Date: Thu, 5 Jul 2018 11:14:28 -0400
Subject: [AstroPy] Parse DETSEC etc..
In-Reply-To: <250df15ff0ed43383d0ac4c9fe8b43dafddabc6f.camel@kubanek.net>
References: <250df15ff0ed43383d0ac4c9fe8b43dafddabc6f.camel@kubanek.net>
Message-ID: <3705E546-C398-4482-ADA2-45CBE7217C61@gmail.com>

the ccdproc package has a utility for doing this: 

http://ccdproc.readthedocs.io/en/latest/api/ccdproc.utils.slices.slice_from_string.html#ccdproc.utils.slices.slice_from_string

it even accounts for FITS-style numbering that starts at index 1. 

tim

> On Jul 4, 2018, at 21:21, Petr Kub?nek <petr at kubanek.net> wrote:
> 
> Hi Astropy,
> 
> I might be stupid, but I don't see any way how to parse DETSEC and
> similar FITS header you can encounter in (multiframe) FITS files. I of
> course know how to parse that with e.g. re and mapping result to float
> array, but is there a better way? See for the code (yes, that will
> become a function..unless I know about better way):
> 
> from astropy.io import fits
> import re
> 
> ff=fits.open('file.fits')
> r = re.compile('\[(\d+):(\d+),(\d+):(\d+)\]')
> g = r.match(ff[1].header['DETSEC'])  # DETSEC is '[20:30,40:50]'
> b=map(float, g.groups())
> print(b)
> [20,30,40,50]
> 
> 
> Thanks
> 
> Petr Kub?nek
> http://rts2.org
> _______________________________________________
> AstroPy mailing list
> AstroPy at python.org
> https://mail.python.org/mailman/listinfo/astropy

?
+-----------------------------+
| T. E. Pickering, Ph.D.      |
|   Asst. Staff Scientist     |
|   MMT Observatory           |
| +1-520-305-9823             |
| te.pickering at gmail.com      |
| tim at mmto.org                |
+-----------------------------+



From hack at stsci.edu  Thu Jul  5 13:49:40 2018
From: hack at stsci.edu (Warren Hack)
Date: Thu, 5 Jul 2018 13:49:40 -0400
Subject: [AstroPy] Inverting SkyCoord spherical offset transformations
Message-ID: <d4e2b263-1f8b-4c8c-7c8a-f51421097ebf@stsci.edu>

Hello,

The astropy documentation provides a very useful example of how to 
compute the offset between 2 SkyCoord positions at 
http://docs.astropy.org/en/stable/coordinates/matchsep.html?highlight=spherical_offsets#separations. 

Specifically, the example for "spherical_offsets_to" in that section of 
documentation shows how to derive a delta RA and delta Dec for those 
source positions.? In my case, though, I already have the delta RA and 
delta Dec from an astrometric solution for that observation and want to 
apply those deltas to the source positions to reproduce the astrometric 
catalog positions.? So, the operation I am really interested in would be 
the inverse of this example where those deltas would be applied to the 
"bright_star" SkyCoord position to reproduce (within machine precision) 
the "faint_galaxy" position in RA and Dec.

It could be that some other functionality in Astropy would be more 
suited to providing the right answer to this problem, but the 
documentation does not provide any indication of how to do this (at 
least not where I could find it).? Any pointers would be appreciated.

Thanks,
Warren Hack

From shupe at ipac.caltech.edu  Thu Jul  5 16:47:06 2018
From: shupe at ipac.caltech.edu (David Shupe)
Date: Thu, 5 Jul 2018 13:47:06 -0700
Subject: [AstroPy] Inverting SkyCoord spherical offset transformations
In-Reply-To: <d4e2b263-1f8b-4c8c-7c8a-f51421097ebf@stsci.edu>
References: <d4e2b263-1f8b-4c8c-7c8a-f51421097ebf@stsci.edu>
Message-ID: <75C2B7ED-2962-42C6-AA0B-6A0B8E2E0BFD@ipac.caltech.edu>

Hi Warren,

Here?s a potential solution using a SkyOffsetFrame, building on the spherical_offsets example you linked to:

>>> from astropy.coordinates import SkyCoord, SkyOffsetFrame, ICRS
>>> bright_star = SkyCoord('8h50m59.75s', '+11d39m22.15s', frame='icrs')
>>> faint_galaxy = SkyCoord('8h50m47.92s', '+11d39m32.74s', frame='icrs')
>>> dra, ddec = bright_star.spherical_offsets_to(faint_galaxy)
>>> star_frame = SkyOffsetFrame(origin=bright_star)
>>> my_galaxy = SkyCoord(lon=dra, lat=ddec, frame=star_frame)
>>> galaxy_coord = my_galaxy.transform_to(ICRS)
>>> galaxy_coord.separation(faint_galaxy)
<Angle 1.5902773407317588e-15 deg>
>>> galaxy_coord.to_string('hmsdms')
'08h50m47.92s +11d39m32.74s?

although with limitations, e.g. ?galaxy_coord.spherical_offsets_to(faint_galaxy)? raises an exception about two non-matching frames.

Regards,
David Shupe

> On Jul 5, 2018, at 10:49 AM, Warren Hack <hack at stsci.edu> wrote:
> 
> Hello,
> 
> The astropy documentation provides a very useful example of how to compute the offset between 2 SkyCoord positions at http://docs.astropy.org/en/stable/coordinates/matchsep.html?highlight=spherical_offsets#separations. 
> Specifically, the example for "spherical_offsets_to" in that section of documentation shows how to derive a delta RA and delta Dec for those source positions.  In my case, though, I already have the delta RA and delta Dec from an astrometric solution for that observation and want to apply those deltas to the source positions to reproduce the astrometric catalog positions.  So, the operation I am really interested in would be the inverse of this example where those deltas would be applied to the "bright_star" SkyCoord position to reproduce (within machine precision) the "faint_galaxy" position in RA and Dec.
> 
> It could be that some other functionality in Astropy would be more suited to providing the right answer to this problem, but the documentation does not provide any indication of how to do this (at least not where I could find it).  Any pointers would be appreciated.
> 
> Thanks,
> Warren Hack
> _______________________________________________
> AstroPy mailing list
> AstroPy at python.org
> https://mail.python.org/mailman/listinfo/astropy


From hack at stsci.edu  Fri Jul  6 14:09:05 2018
From: hack at stsci.edu (Warren Hack)
Date: Fri, 6 Jul 2018 14:09:05 -0400
Subject: [AstroPy] Inverting SkyCoord spherical offset, transformations
Message-ID: <461d84e9-f167-76ba-7d3b-0d1e7446a5a5@stsci.edu>

David,

Thank you!? That solution allowed me to apply the offsets I have to the 
test images and align them (better) to the astrometric catalog positions.

The real issue in my opinion would be the non-intuitive nature of the 
required solution, especially for someone not actively developing the 
astropy coordinates code.? I would suggest to the astropy developers 
that they therefore include this solution, or whatever may improve upon 
it (if possible), in the documentation. It would serve as a follow-up to 
the example already in the docs which served as the starting point for 
this question, and show how to 'round-trip' the information being computed.

For now, though, my problem has been solved.

Thanks again!
Warren

From bkent at nrao.edu  Mon Jul  9 17:07:26 2018
From: bkent at nrao.edu (Brian R. Kent)
Date: Mon, 9 Jul 2018 17:07:26 -0400
Subject: [AstroPy] Invitation to submit to the PASP data visualization issue
Message-ID: <4f1beb3d-4368-a8e2-fc21-1f79c1ac51b2@nrao.edu>

The Publications of the Astronomical Society of the Pacific (PASP) is 
producing a **second** focus issue concerning the topic of Data 
Visualization in Astronomy.? Submissions will be handled through the 
normal PASP referee process with regular page charges.? You can read our 
first well-received focus issue and its intro video here:

http://iopscience.iop.org/journal/1538-3873/page/Techniques-and-Methods-for-Astrophysical-Data-Visualization

The editors feel that this focus issue would provide a great practical 
benefit in continuing the success of tool/technique/code sharing 
generated from the publication of our first focus issue on data 
visualization.? This new focus issue will publish science results or 
methods discovered with innovative tool/package descriptions.? All 
articles will have a shared theme under the focus of data 
visualization.? Submissions could address one or more of the following 
topics and/or questions:

- What useful tool have you built/utilized for astronomy?
- What science problem has your tool/package/software/technique solved?
- What data visualization technique can be shared with the astronomical 
community?

If your submission would benefit from popular visualization tools like 
Oculus Rift, HTC Vive, or Google Cardboard, tell us - we would enjoy 
having those types of articles.? We also strongly encourage authors to 
share their code via repositories including GitHub and the Astrophysics 
Source Code Library (ASCL).? In addition, we would like to have authors 
create video abstracts showcasing their visualization work.

We would like to receive all submissions by the end of the calendar year 
2018.
If you are interested in submitting a paper, please let use know in the 
next few weeks.? We're happy to discuss your submission ideas. If you 
know of other members of the astronomical community whose recent work 
would be appropriate
for this focus issue, please send us their name/contact information so 
that we may extend the invitation
to them as well.

Sincerely,
Brian R. Kent, NRAO
PASP Focus Issue Guest Editor
bkent at nrao.edu

Jeff Mangum, NRAO
PASP Editor-in-chief
jmangum at nrao.edu