Julia is most mature in what areas?

  • Thread starter Thread starter deltapapazulu
  • Start date Start date
  • Tags Tags
    Areas
AI Thread Summary
The discussion revolves around the Julia programming language, focusing on its maturity in ecosystem, libraries, and industry adoption. Participants express curiosity about Julia's applications, particularly in computational physics and machine learning, noting its growing popularity as a competitor to Matlab. While Julia's community is robust, it is smaller than those of Python and JavaScript, which can limit resources for new learners. Key strengths of Julia include its performance with graphics processing units and its suitability for machine learning, although it faces challenges such as longer startup times due to code recompilation and a less developed integrated development environment (IDE) compared to Matlab. The conversation also highlights Julia's rankings in programming language indexes, with mentions of its position in the IEEE Spectrum and TIOBE rankings. Participants share personal experiences with Julia and other languages, emphasizing a broader interest in programming and the exploration of new technologies.
deltapapazulu
Messages
84
Reaction score
13
Hello,

up late toying with Julia programming language. Just curious if anyone here has knowledge on something. In which areas has the Julia programming language gained the most maturity in terms of ecosystem, libraries and practical industry adoption?
 
Technology news on Phys.org
sysprog said:
This is a good casual article that discusses Julia compared to Python:
https://devathon.com/blog/julia-vs-python-which-programming-language-is-better/
thanks, yes that is a good article. I'm trying to find some avenue into the language besides just the obvious of documentation and basics tutorials (which I am doing). My highest math knowledge is Calc based Linear Algebra and ODE. Which is not great and not really high enough for some things I'd like to dabble with. But I'd still like to find some project to do with Julia. I'm the chap who started the javascript Latin nouns thread. I'm still learning that, but I got some good work at my job recently so I don't have to worry so much about landing something in programming right away. I was desperate and out of work at the time.

Anyway, I have been learning RUST mainly but also some Julia over the last few weeks and am thoroughly enjoying it and now that I've got some decent cash flow again I'd love to focus more on these two languages. Learning Javascript was an outstanding primer for me and wet my whistle for the more strictly programming side of things. RUST has really been my main focus and will continue to be. This Julia excursion is mainly just me doing some exploring.
 
Nice to learn that things are going better for you at work, @deltapapazulu. Regarding Julia, maybe you should look at https://julialang.org/community/. Although the Julia community is not as large as the Python or Javascript communities, it's very robust, and has a wealth of resources available.
 
sysprog said:
Nice to learn that things are going better for you at work, @deltapapazulu. Regarding Julia, maybe you should look at https://julialang.org/community/. Although the Julia community is not as large as the Python or Javascript communities, it's very robust, and has a wealth of resources available.

Yeah, thanks, I am kind of surprised Julia is not more discussed in Physics Forums. I figured I'd start a thread to see if there where any Julia buffs around here. Anyway here is a good discussion regarding Julia's adeptness for computational physics. Some of the better commentary is further down the thread.

Is Julia suitable for computational physics?
 
I’ve brought up Julia several times. It’s new and has been gaining a lot of traction. I think it’s a serious competitor to Matlab although Matlab maturity and vast proprietary libraries clearly has the edge.

it seems that Machine Learning is the big area where Julia is growing of course a lot of lower level libraries are coming online too but they are indirectly in the service of machine learning.

one strength it has is in its use of graphics boards through one of its parallel Izard on libraries Which Matlab hasn’t fully exploited yet Throughout their product line.

A major weak point in my opinion is its need to recompile code at startup which makes it appear slower than Matlab. In particular, the first thing I tried was a simple plot and it took much longer for it to popuo than I expected and as compared to Matlab.

the other major weakness is an IDE like Matlab. Of course they have Jupyter notebook IDE and Juno editor but they don't compare with the Matlab IDE which makes it difficult to attract Matlab folks to the new language.
 
  • Like
Likes deltapapazulu
In the enterprise category now at # 13 at IEEE Spectrum Top Programming Languages:

IEEE_Position_of_Julia.JPG
 
  • Like
Likes deltapapazulu and sysprog
In my opinion, every programmer would benefit from checking out rosettacode.org ##-## it's a real eye-opener.
 
  • Like
Likes jedishrfu and deltapapazulu
  • #10
I watched vid by somebody who convinced me that spectrum.ieee.org's ranking is somewhat more reliable and reflective of reality than tiobe (but I do visit tiobe a lot). I didn't mean to be misleading, the ranking I posted screen of was for ENTERPRISE category only. Here is ranking in all 4 cats, Web, Enterprise, Mobile, Embedded. But even then it has Julia at #20 and RUST at #17. (see below)
My main focus right now is RUST, I'm pretty sold on it becoming more and more bullish in next 2 years and I kind of want to be there for it if that happens. Big 'if' of course. I learned C++ through beginners and some intermediate a few years ago so not completely new to programming but doing a real project in Javascript (last month) made me want to get into some closer to metal programming. I realize that is not what Julia is, I'm just exploring it due to the buzz around it.

https://spectrum.ieee.org/top-programming-languages/
 
  • #11
just thought I would add this to this thread.

I managed to get R, Python, and Julia set up in Jupyter Notebook on Win10 without Anaconda. Here is determinant of a 4x4 matrix in all three languages. Notice the simpler format in JULIA:

Julia_Python_R_comparison.png
 
  • Like
Likes sysprog
Back
Top