From perry at stsci.edu  Thu Jul 14 10:56:14 2005
From: perry at stsci.edu (Perry Greenfield)
Date: Thu, 14 Jul 2005 10:56:14 -0400
Subject: [AstroPy] Re: [SciPy-user] pyfits question: handling gzipped fits
	files
In-Reply-To: <rowen-9B11C4.07384014072005@sea.gmane.org>
References: <rowen-9B11C4.07384014072005@sea.gmane.org>
Message-ID: <a752de03d75fcec9e0902536b0f07564@stsci.edu>

Hi Russell,

The astropy list is probably a better forum for this.

On Jul 14, 2005, at 10:38 AM, Russell E. Owen wrote:

> Is there some reasonable way to get pyfits to read a gzipped fits file
> (in a way that works on Windows and unix)?
>
> I realize I can unzip the file first, but I was hoping to have it
> directly read (unzipped on the fly).
>
> I can't see any command that takes a file-like object (instead of a 
> file
> path), so I can't see how to take advantage of Python's gzip module.
>
> I realize I can just edit the code and add this in, but, well...if
> there's a better way I'd love to know about it and if there isn't, I'd
> like to request that it be modified to handle gzipped files
> automatically.

Not yet, but after version 1 is out, it's one of the first things we'd 
like to add (as well as support for the compression conventions that 
CFITSIO supports). We don't have time right now to do it (though it 
shouldn't be hard at all). If you or someone would like to add it, we 
would welcome such a patch.

Perry



From mpfitz at berkeley.edu  Fri Jul 15 07:41:33 2005
From: mpfitz at berkeley.edu (Michael Fitzgerald)
Date: Fri, 15 Jul 2005 04:41:33 -0700
Subject: [AstroPy] Re: [SciPy-user] pyfits question: handling gzipped fits
	files
In-Reply-To: <a752de03d75fcec9e0902536b0f07564@stsci.edu>
References: <rowen-9B11C4.07384014072005@sea.gmane.org>
	<a752de03d75fcec9e0902536b0f07564@stsci.edu>
Message-ID: <200507150441.37510.mpfitz@berkeley.edu>

On Thursday 14 July 2005 07:56, Perry Greenfield wrote:
> On Jul 14, 2005, at 10:38 AM, Russell E. Owen wrote:
> > Is there some reasonable way to get pyfits to read a gzipped fits file
> > (in a way that works on Windows and unix)?
<snip>
> Not yet, but after version 1 is out, it's one of the first things we'd
> like to add (as well as support for the compression conventions that
> CFITSIO supports). We don't have time right now to do it (though it
> shouldn't be hard at all). If you or someone would like to add it, we
> would welcome such a patch.

Hello,

Here is a patch against pyfits 0.9 that does the trick for me on unix.  It 
simply uses the gzip.open() function to open the file instead of 
__builtin__.open() if the filename ends in .gz.

Unfortunately, the patch is broken in the most recent version of pyfits 
(0.9.8.2).  I don't know in what version the break first occurs, but the 
problem must've started when pyfits began using the two-argument form of File 
objects' seek() method.  The (optional) second argument is the 'whence' 
direction flag.  In short, the gzip.GzipFile class does not share the 
interface of the File class, and can't be used as a drop-in replacement in 
the most recent version of pyfits.

For the sake of consistency, it would be ideal if the File/GzipFile interfaces 
were compatible, which means this should be fixed in Python's library.  I 
haven't poked around much for the best way to achieve this.  I did notice, 
however, the bz2 module's BZ2File class does implement the two-argument 
seek() method.  Perhaps the update to the gzip module can be modeled off the 
bz2 module's implementation.

Finally, since I can't readily find pyfits 0.9 on the web, I've temporarily 
put the patched 0.9 version here:

	http://astron.berkeley.edu/~fitz/pyfits.py

Mike

-------------- next part --------------
A non-text attachment was scrubbed...
Name: pyfits.py.diff
Type: text/x-diff
Size: 966 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/astropy/attachments/20050715/0eaf9d12/attachment.diff>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
URL: <http://mail.python.org/pipermail/astropy/attachments/20050715/0eaf9d12/attachment.sig>

From stephen.walton at csun.edu  Sun Jul 17 15:20:22 2005
From: stephen.walton at csun.edu (Stephen Walton)
Date: Sun, 17 Jul 2005 12:20:22 -0700
Subject: [AstroPy] Re: [SciPy-user] pyfits question: handling gzipped
	fits	files
In-Reply-To: <200507150441.37510.mpfitz@berkeley.edu>
References: <rowen-9B11C4.07384014072005@sea.gmane.org>	<a752de03d75fcec9e0902536b0f07564@stsci.edu>
	<200507150441.37510.mpfitz@berkeley.edu>
Message-ID: <42DAAF76.4030503@csun.edu>

Does Pyfits support the internally compressed FITS files described in 
this link

http://ledas-www.star.le.ac.uk/lheasoft/fitsio/compression/compress_image.html

of which I see Perry is a co-author?  Of course, reading such files is 
one thing, having software to write them quite another.  CFITSIO can do 
it, I think.




From perry at stsci.edu  Mon Jul 18 09:18:32 2005
From: perry at stsci.edu (Perry Greenfield)
Date: Mon, 18 Jul 2005 09:18:32 -0400
Subject: [AstroPy] Re: [SciPy-user] pyfits question: handling gzipped
	fits	files
In-Reply-To: <42DAAF76.4030503@csun.edu>
References: <rowen-9B11C4.07384014072005@sea.gmane.org>	<a752de03d75fcec9e0902536b0f07564@stsci.edu>
	<200507150441.37510.mpfitz@berkeley.edu>
	<42DAAF76.4030503@csun.edu>
Message-ID: <bff1e789522576db99a8a96f93ddd412@stsci.edu>


On Jul 17, 2005, at 3:20 PM, Stephen Walton wrote:

> Does Pyfits support the internally compressed FITS files described in  
> this link
>
> http://ledas-www.star.le.ac.uk/lheasoft/fitsio/compression/ 
> compress_image.html
>
> of which I see Perry is a co-author?  Of course, reading such files is  
> one thing, having software to write them quite another.  CFITSIO can  
> do it, I think.
>
Not yet, but should be within a 2-3 months (resources permitting).

Perry



From rowen at u.washington.edu  Fri Jul 22 17:14:21 2005
From: rowen at u.washington.edu (Russell E Owen)
Date: Fri, 22 Jul 2005 14:14:21 -0700
Subject: [AstroPy] PyGuide 2.0 released
Message-ID: <p06230903bf070ffdc6e5@[128.95.99.44]>

PyGuide 2.0 is now available from 
<http://www.astro.washington.edu/rowen/PyGuide%20Manual.html>.

PyGuide offers image analysis routines for telescope guiding, 
including a star finder, centroider and star shape fitter. The 
algorithms are designed to work well with partial images, e.g. where 
much of the image of a star is lost down a spectrograph slit or off 
the edge of a coherent fiber optic bundle.

Version 2.0 is a major overhaul based on use at Apache Point 
Observatory. In particular, centroid and starShape have been much 
improved.

StarShape still does not measure ellipticity. If anyone has a 
reliable algorithm for measuring ellipticity even on heavily masked 
data, I would like to know about it. By reliable I mean that it 
includes an error estimate so I know when not to trust the results.

-- Russell

P.S. Future plans:
- Add a measure of ellipticity
- Replace satLevel with a saturation mask
- Work with Float32 images instead of UInt16
the last two changes will make PyGuide easier to use with flat-fielded images.


From perry at stsci.edu  Tue Jul 26 17:52:02 2005
From: perry at stsci.edu (Perry Greenfield)
Date: Tue, 26 Jul 2005 17:52:02 -0400
Subject: [AstroPy] "Using Python for Interactive Data Analysis" tutorial
	available
Message-ID: <6f83e4841620a3e3466ad71c5084076f@stsci.edu>

A tutorial that illustrates how Python can be used to analyze and 
visualize astronomical data (much like one can do in IDL) has been 
written and now is available at:

http://www.scipy.org/wikis/topical_software/Tutorial