Julia 101

About myself

  • Georg Wechslberger (PhD)
  • Working at Technical University of Munich
  • Research Area: Numerical Analysis
  • Research Topic:
    • Contour Optimatisation
    • Riemann-Hilbert Problems
  • ... or "Finding the shortest path to free beer."

Julia: History

  • development started in 2009 by Jeff Bezanson, Stefan Karpinski, Viral B. Shah, and Alan Edelman
  • public annoucment on 14.02.2012
  • first release v0.1 on 14.02.2013
  • current release v0.7
  • v1.0 expected to be released end 2017 / beginning 2018

Goal: "being greedy"

  • speed of good C code
  • usability of Python
  • Matlab like syntax
  • string handling tied in as in Perl
  • as easy as R for statistics
  • simple and interactive

Why should I use julia?

open source

  • usable free of charge
  • MIT License
  • growing community

In [1]:
Pkg.add("Requests")
using Requests


INFO: Package Requests is already installed
INFO: METADATA is out-of-date — you may not have the latest version of Requests
INFO: Use `Pkg.update()` to get the latest versions of your packages

active development


In [12]:
img = get("https://pkg.julialang.org/img/allver.svg");
display("image/svg+xml",String(img.data))


fast

But ... but ... all my existing code?

  • Matlab? -> similar syntax, just run it (with some modifications)
  • C / Fortran ? -> can be directly called, using build in interface
  • Python ? -> can be directly called, using PyCall interface
  • Mathematica ? -> can be directly called, using Mathematica.jl interface

Why shouldn't you use Julia?

really active development

  • breaking changes with every major release (might get better after v1.0
  • many packages are still in development
  • "the" package for a commmon usage pattern can change
  • package development may stop / switch to another approach

In [ ]: