Learning Python — The DIY Way

by Guest on July 17, 2012

This is a guest post from Prof. Marla Geha at Yale University. Her research focuses on the formation, evolution, and destruction of dwarf galaxies.

I’ve wanted to jump on the python bandwagon for a while. I’m an IDL programmer, but IDL licenses are becoming more annoying to maintain/afford. The excuse that python doesn’t provide the same astro-specific functionalities doesn’t seem valid anymore.  I was further motivated to learn python as I teach an undergraduate astronomy research methods class — teaching in python (versus IDL or IRAF) just makes more sense.

With a short attention span and endless distractions, there is little chance I was going to ‘sit down’ for a few days on my own to work through a python book or website. Instead I decided to gather a number of like-minded people in my department and run a do-it-yourself (DIY) python workshop. We worked one full day and 2 half days. Everyone in our group was an IDL user. No python expert was needed, although we did bring one in on the second day for questions.

Our workshop was successful and I’ve written up an AstroBetter wiki page so others can follow, linked off the main Python wiki page. There are an overwhelming number of python resources on the web and the wiki page details the resources that we felt work best in this context. This includes video tutorials, lectures, and plenty of hands-on exercises. With only a 2.5 day investment and plenty of coffee, I’m a python convert.

{ 8 comments… read them below or add one }

1 John O'Meara July 17, 2012 at 2:28 pm

“IDL licenses are becoming more annoying to maintain/afford”. This argument alone makes me want to jump in on the python bandwagon. Gonna spend a day with that wiki and see where it takes me. Thanks Marla!

Reply

2 Jim Davenport July 19, 2012 at 11:24 am

Agreed… since IDL 7 is already a PITA to run on Macs, we’d probably all be wise to switch.

3 Sarah Tuttle July 17, 2012 at 9:59 pm

Eilat told me about this, and her enthusiasm plus the wikipage makes me want to be a copycat – thanks for the inspiration. I think I’m going to rally the troops at UT.

Reply

4 Jane Rigby July 24, 2012 at 10:21 am

Marla, thanks for doing this! I’m inspired to copycat your DIY workshop.

Reply

5 Claire Lykou July 24, 2012 at 10:51 am

Looking forward to test this! Thanks for posting. I’d also like at some point to stop using commercial programs that I cannot use outside the Uni.

Reply

6 JP July 24, 2012 at 11:39 am

Thanks. Some interesting links to resources from the DIY workshop and the wiki.

Reply

7 Jane Rigby July 30, 2012 at 8:29 am

This is really great! Thanks to Marla for making the agenda/syllabus public, which makes it easy for others to copycat the workshop! We’re holding one at my institution next week, based on Marla’s notes plus the Austin SciPy conference. More than 30 scientists showed up last week for the first meeting of our new Python users group!

Reply

8 Richard Scalzo August 7, 2012 at 8:36 pm

Wow, I didn’t even know there was a SciPy conference.

I’ve never looked back ever since I ditched C++ for Python, and as others have said, the lack of licenses makes Python much more appealing than IDL — especially for high-throughput pipeline work where you may have many processes running at once on different chunks of data. You won’t believe how much money they blew on IDL licenses at LBL.

Thanks also for the links — looks like a lot of good stuff in there!

Reply

Leave a Comment

Previous post:

Next post: