4 ms·
Hey Chris! Thanks for all the hard work btw. I had a bug originally with the Julia Opt packages. I think the "Cbc" and "Clp" packages had issues on 0.7 & 1.0 bu
by 4thaccount 8y ago
Hey Chris! Thanks for all the hard work btw. I had a bug originally with the Julia Opt packages. I think the "Cbc" and "Clp" packages had issues on 0.7 & 1.0 builds. This one was relatively easy to solve in that the windows path it was looking for didn't match the directory it was installed in. I actually had trouble with this one on Ubuntu as well.
Just yesterday I tried to install "Plots" on Windows7 using Julia 1.0 and the installer popped out a ton of red errors. I'll have to go look again later today, but they didn't seem as straight forward to fix. I doubt that helps a lot, but I'm curious why these sorts of things seem to happen a lot (or at least in my limited experience). When the package manager works, it is fantastic!
I'd like to use Julia for optimization, sparse matrix math, genetic algorithms, graph theory, and plotting. Having a free and fast tool to do all of this is pretty amazing. GNU Octave is cool, but doesn't seem to have most of what I need. Matlab and Mathematica are powerful, but unreasonabley expensive if you don't need to use them daily. They do have extremely good integration of scientific tools like plotting. Mathematica is an amazing tool btw and I hope Julia gets as many features some day.
- ChrisRackauckas 8y agoAhh yes, optimizers with binaries can sometimes be a pain. Plots should be good though: I don't experience issues with it. Every once in awhile GR's download server goes down, so `]build GR` or `]build Plots` can help re-download the binary.
- 4thaccount 8y agoFinally got "Plots" to work. Not sure what I was doing wrong the first time. It takes a long time to run though...maybe some kind of first pass compilation? The Lorenz Attractor demo just complained about not being able to connect to "GKS socket application", but it seems to have written out a .gif that runs in IE and is seriously cool :)