from __future__ import division
import os
import sys
import glob
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
%matplotlib inline
%precision 4
plt.style.use('ggplot')

Julia and Python¶

You will need to install Julia from http://julialang.org/downloads/ and make sure it is on your path. On Ubuntu you can install via apt-get with

sudo add-apt-repository ppa:staticfloat/juliareleases
sudo add-apt-repository ppa:staticfloat/julia-deps
sudo apt-get update
sudo apt-get install julia

Within an instance of Julia, run the following commands:

Pkg.update()
Pkg.add("PyCall")
Pkg.add("IJulia")

Then finally install pyjulia from https://github.com/JuliaLang/pyjulia

Make sure that you can start julia line - if not, add a symlink to it.

%install_ext https://raw.githubusercontent.com/JuliaLang/pyjulia/master/julia/magic.py

Installed magic.py. To use it, type:
  %load_ext magic

%load_ext magic
%julia @pyimport matplotlib.pyplot as plt
%julia @pyimport numpy as np
%julia @pyimport numpy.random as npr

Initializing Julia interpreter. This may take some time...

Defining a function in Julia¶

%%julia

function fib(n)
    a, b = 0.0,  1.0
    for i = 1:n
        a, b = a+b, a
    end
    return a
end

<PyCall.jlwrap fib>

Using it in Python¶

jfib = %julia fib

jfib(100)

354224848179261997056.0000

Using Python libraries in Julia¶

%%julia

xs = npr.multivariate_normal([0,0], np.eye(2), 100)
plt.scatter(xs[:,1], xs[:, 2], s=30);

<matplotlib.collections.PathCollection at 0x1220f3a50>

Benchmarking¶

%timeit jfib(100)

10000 loops, best of 3: 22.9 µs per loop