[AstroPy] Proliferating py-astro-libs
Jaime Perea
jaime at iaa.es
Tue Jun 14 17:37:58 EDT 2011
Hello list
My first post to this list.
I have been working for a while now in a rather complicated data reduction pipeline. I don't considerer myself as an
expert, no way, but it's a long time ago when I wrote my first python script. I firmly agree with the idea of an
astropy meeting, I guess that there is lot of people here in the old Europe that will enjoy meeting other people and
share knowledge.
In some sense I think that python is just like a lego, there are basic pieces, bricks that put together allow the
programmer to write smart code. With this idea I would like to you to help me to identify such bricks, some of them are
easy, some maybe not.
- Numpy / SciPy - obvious,
- Matplotlib for the 2D, there exist some other alternatives such as plplot.
- pyfits - This is a master piece of software I still remember something like
open(unit=lun,file=fits,form='unformatted',recl=2880) and I don't want to see it again.
- *wcs, for this time of big catalogs, they are also mandatory.
- pyraf - this allow to access tons of astronomy sofware already there
With such building blocks you can do a lot of things. But I still would like to have an interface to other stuff that
we use a lot:
- Statistics - something handful, easy, scipy contains a lot, but there are astronomical tasks that are so repetive that
perhaps may deserve their own building block.
- Access to databases and eventually Virtual Observatory. e.g. For calibration we access 2MASS very often, we use
urllib, but I would like to have something with a more general scope.
- Also I tend to think at short wavelengths, but I would love to identify to what extent software such as AIPS or GIPSY
is pythonic enough now, and if we can assume that there exist interfaces to such soft.
- number crunching - with stuff such as numpy anv even going further to mpi4py or pypar or whatever it turns possible
to write numeric (modeling) code. At least for the analysis stage.
- interactive data analysis, can we turn ipython or whatever "more astronomical" or do we have to do something
different?
I think we need to identify such pieces and write glue code that allow us to interact with them. So perhaps the first
task has to be an inventory of sofware and human resources. And if this is done perhaps we may realize the way
we have to manage the Project.
I support the idea of a meeting in fall.
Receive all my best
--
Jaime D. Perea Duarte. <jaime at iaa dot es>
Linux registered user #10472
Dep. Astrofisica Extragalactica.
Instituto de Astrofisica de Andalucia (CSIC)
Apdo. 3004, 18080 Granada, Spain.
More information about the AstroPy
mailing list