Wednesday, April 25, 2012

Julia


Julia is a new "high-level, high-performance dynamic programming language for technical computing." Website here and rationale here. Interview with Stefan Karpinski.

I followed the instructions here to download and build it. The only issue is that I failed to read the platform-specific notes before starting, so I had the wrong (old) gfortran. There are links in the docs, but I followed the notes in this thread and got it from (download link).

[ UPDATE: re the first comment, some discussion here. ]

4 comments:

Mao Jianfeng said...

Thank you for introducing such an attractive language. I would like to know if there is any case of Julia used for bioinformatic studies. Have you experienced any?

Thanks

telliott99 said...

No, I don't know any specific examples. I posted to give it some buzz, but I doubt you'll get many R users to switch. For me, I like Python's clarity a lot, and would just add C or C++ code when needed, whether by using subprocess.Popen, or Cython, or ctypes, or even just numpy.

Diego Zea said...

Pypy is almost as fast than Julia ;) Try it [ Numpypy is going to be more faster ? ]

Diego Zea said...

WOW!!! I read the documentation of Julia... The language it's amazing and have a lot of useful things borrowed from other languages. A clear evolution. I believe in Julia's future ;)

But, It's truth... People doing science, coming from statistics, biology, physics are more likely to used always the same old-language. Have a big community (maybe) it's going to take a while...