What software do you use for coding and/or graphics?

[kgriffin]

I know this was always a topic of debate back in my days at Albany, and here at Wisconsin I note many of the same divisions, albeit everyone here is a bit less competitive

I know many places teach undergrads GEMPAK in their synoptic meteorology class, but I know some also use other programs.

Which one of the above choices do you use (even a little bit), and what do you love (and hate!) about it?


Personally, I was teethed on GEMPAK and used it for many years until I wanted a bit more control over my graphics and to make some more complex data plots from similarly complex statistical analysis.

Although I took a class where MATLAB was used for a lot of statistical analysis, I've come to be a big fan of NCL. It has a pretty steep learning curve and is by no means the fastest of programs, but you have super fine control over just about anything you do. It also has a lot of more complex functions built in and can be somewhat time saving in that regard. I know it's my first choice these days!

[Derek Mallia]

I generally use NCL, R and python. 

R does a really nice job with calculations and statistics which is always very useful in the science world. It also has a fairly easy to use netcdf package that can read in or write out netcdf. Its also really good when dealing with text data.

In terms of plotting, R isn't exactly the prettiest language to do this with its limited options. Generally this is when I use NCL since it can make some terrific looking plots/maps. 

I use python for everything else!

I try to avoid matlab and IDL as much as possible as there is always the possibility that I won't have access to these languages in the future, hence I rather master the free programs which I can carry with me wherever I go!

[vanna]

If I could vote for Python multiple times I would.  I've used Frotran, Grads, Matlab, IDL and NCL, but Python is by far my favorite language and the most versatile.  I did not realize everything that Python was capable of doing when I started using it, but after some exploration into some of the packages and visiting Scipy this summer, it really is a great language. There are packages that run through C and Fortran, so the code can run just as quickly on machines with C or Fortran or in some cases still run (slower) on machines with only Python.  The graphing packages are easy to use out of the box or with a bit more work can do anything.  And the Pandas package in Python is spectacular for statistical uses and dealing with irregular data sets.

[Matt Sienkiewicz]

MATLAB is currently my favorite since it seems to be excellent for statistical analysis of time series data, which is all my life has become lately. I think I currently have three separate licenses on three different machines. I also really enjoy its netCDF functionality. 

NCL is definitely high on my list of languages to learn.

I think my favorite thing about this poll is that RIP (Read/Interpolate/Plot) is not listed.

[Derek Mallia]

Yea quite a few people I know use RIP with their WRF graphics. Makes me cringe though I am a bit of a graphics snob