3/31 - Julia - STO 442 - 5:00-7:00

[wkearn]

Alexey and I were chatting the other day, and he said there might be some interest in an intro to Julia. I'd love to drop by one week and show it off.

I could imagine a tutorial getting people programming in Julia, or I could do more of a demonstration of how Julia's design enables writing very fast code fairly easily. The former might be more useful and easier to follow, but Julia really shines when you're neck-deep in numerical computations, and you need a fairly complex application before you can really see the benefits. Let me know what you all would be interested in.

And now that I've found this repo, I'll start coming to the study group. Looking forward to meeting everyone!

[ceholden]

Awesome! I am very excited to learn more about Julia. I've heard much about the JIT magic that makes it fast and that Julia can basically sit on top of the existing scientific Python and C/Fortran ecosystems, but I haven't seen many demonstrations that weren't Fibonacci level of trivial. Personally, I would favor seeing project examples of yours using Julia that get more in depth and, as you say, demonstrate the strengths of the language.

[tomhohenstein]

+1 @wkearn - any interest in scheduling a day / time to talk about Julia?

[wkearn]

Sure! Looking at #19, it seems like March 31 is free. That works for me.

[tomhohenstein]

great! I'll start doing the updates

[wkearn]

I've uploaded a little Julia module that we'll be playing around with tomorrow to wkearn/GeometryRandomFields.jl. There is a Jupyter notebook that we'll be looking at in the examples directory.

If you have a Google email account, you can experiment tomorrow with Julia on JuliaBox. It's a little tricky to get everything properly set up, so I'll go over that tomorrow. If you really want to try it out now, let me know, and I'll help you through it.

If you already have Julia installed locally or want to install it locally, you'll need to make sure that you have version 0.4 of Julia (the current release, not the development v0.5). Everything should work with any official release of v0.4, but JuliaBox will be using the most recent v0.4.5.

Once you have Julia installed, fire up what's called the REPL (in a terminal, type julia) and enter the following commands to get the packages we'll need to use:

> Pkg.add("Plots")
> Pkg.add("GR")
> Pkg.add("IJulia")

This adds Plots.jl and the GR.jl plotting backend as well as the IJulia kernel which connects Julia to Jupyter notebooks. You can also use Plots.jl with PyPlot.jl (Pkg.add("PyPlot")), though the beauty of Plots.jl is that everything should work the same regardless of which plotting backend you use.

You should, at this point, be able to run through the notebook in the examples folder of the GeometryRandomFields.jl package I linked to without any errors. Let me know if anything goes wrong.

I can open a pull request to add the notebook to the lessons folder, or if we figure out how to use submodules (#21), I can add the GeometryRandomFields.jl repo as a submodule. Let me know which is a better option.

[tomhohenstein]

@wkearn great work! I'm really looking forward to tomorrow's session.

[tomhohenstein]

I counted 7 people - great session

[mybeliza]

Hello All,

A friend of mine bought two tix to the Questrom spring formal tomorrow. Now, she has injured her foot and can't make it. Anybody looking for spring formal tix? I know they've been sold out for a while now. She's willing to sell them at $80 instead of $90 for two.

Regards, Myrdell

On Tue, Apr 5, 2016 at 4:45 PM, Tom Hohenstein notifications@github.com wrote:

I counted 7 people - great session

— You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub https://github.com/bulib/studyGroup/issues/37#issuecomment-205980062