<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
I very much agree with what marshall said. I am really in favour of
a modular design rather than these monolithic packages. For example:
I use pyfits for my fits files, it can't do coordinate
transformation and it can't do light curve fitting, but it's fast
and good at what it does. In addition, it puts the data in a format
that other packages understand: numpy arrays. The same with ephem,
pywcs, aplpy, atpy and others (I'm sorry if I didn't mention your
package). I think we need to build an ecosystem of these small
tools, that are bug free and easy to use before going to big
packages (Sage is a good example of this working). One big point is
that they need to be interoperable. <br>
I'm going to reiterate what I said on this topic before: we need
standard base classes so that if someone writes a new tool it can
easily interact with other tools. <br>
Andy Casey and me are currently working on a base class for 1d
spectra. It's very easy to use and very simple. We will release it
into the wild in a month or two and hope that this will bring out
the discussions about APIs for base classes. <br>
<br>
Cheers<br>
Wolfgang<br>
<br>
On 10/06/11 8:25 AM, Marshall Perrin wrote:
<blockquote
cite="mid:C0AECAED-9701-4A76-B225-CA2716CD8430@stsci.edu"
type="cite">
<meta http-equiv="Content-Type" content="text/html;
charset=ISO-8859-1">
<div>On Jun 9, 2011, at 12:54 PM, Stefan Czesla wrote:</div>
<div>
<blockquote type="cite">
<div>Dear all,<br>
<br>
we would like to let you know about our recent release of a
-- hopefully --<br>
useful contribution to Python's astronomy community, namely,
our PyAstronomy<br>
package (yes, there have been more inspired names...). It
consists of<br>
several subpackages providing the following functionality:<br>
<br>
- funcFit: An XSPEC like fitting interface for scipy and
pymc<br>
- modelSuite: Currently, models for transit light-curves and
Rossiter McLaughlin effect<br>
(calculating and fitting)<br>
- pyTiming: A collection of periodograms<br>
- pyasl: Our tiny contribution to porting the IDL AstroLib
to Python.<br>
</div>
</blockquote>
<div><br>
</div>
<div><br>
</div>
<div>Hopefully without sounding too critical of you in
particular, I'm going to ask: do we as a community really need
<i>yet another</i> separate python library for astronomy and
yet another attempt at building a core set of routines ported
from the IDL library? </div>
<div><br>
</div>
<div>There is now astrolib (<a moz-do-not-send="true"
href="https://trac6.assembla.com/astrolib">https://trac6.assembla.com/astrolib</a>
), also pyAstroLib (which appears to be something entirely
different than astrolib? <a moz-do-not-send="true"
href="http://sourceforge.net/projects/pyastrolib/develop">http://sourceforge.net/projects/pyastrolib/develop</a>),
astropysics (<a moz-do-not-send="true"
href="http://packages.python.org/Astropysics/">http://packages.python.org/Astropysics/</a>)
, now your pyAstronomy, and probably more. It looks like
PyAstronomy contains wavelength conversions (already
implemented in pyastrolib and astropysics) and coordinate
conversions and precession (already in coords, astropysics,
pySLALIB, & more). </div>
<div><br>
</div>
<div>I'm going to be provocative here: <i>As a community, we
are doing something wrong
</i>if everyone wants to start their own new module rather
than contributing to a common shared open-source core. We are
clearly doing something wrong if people repeatedly implement
the same basic functions rather than building on what's
already there. What do we need to do differently? How can we
make it easier to use a shared repository and shared namespace
for all this?</div>
<div><br>
</div>
<div>I'm not arguing that somehow there has to be one true
library that everyone uses, one ring to rule them all. But
wouldn't it be better if there were a more straightforward
choice that community members could easily contribute to,
leading to a more coherent overall system? I would assert that
astrolib is probably the best candidate for growing into this,
but I don't really care which one the community rallies
behind. I'd just like to have some sense that we might
actually be converging towards an IDL-like unified core rather
than just proliferating little packages. Currently we're
just forming dwarf galaxies; how can we get them to accrete
together to build a grand design spiral? </div>
<div><br>
</div>
<div> - Marshall</div>
<div><br>
</div>
<div><br>
</div>
<div><br>
</div>
</div>
<pre wrap="">
<fieldset class="mimeAttachmentHeader"></fieldset>
_______________________________________________
AstroPy mailing list
<a class="moz-txt-link-abbreviated" href="mailto:AstroPy@scipy.org">AstroPy@scipy.org</a>
<a class="moz-txt-link-freetext" href="http://mail.scipy.org/mailman/listinfo/astropy">http://mail.scipy.org/mailman/listinfo/astropy</a>
</pre>
</blockquote>
<br>
</body>
</html>