A Beginner's Look at Julia

Over the past month or so, I’ve been playing with a new scientific programming language called ‘Julia’, which aims to be a high-level language with performance approaching that of C. With that goal in mind, Julia could be a replacement for the ‘multi-language’ problem of needing to move between R, Python, MATLAB, C, Fortran, Scala, etc. within a single scientific programming project.  Here are some observations that might be helpful for others looking to get started with Julia.

Get used to ‘Git’ and ‘make’

While there are pre-built binaries for Julia, due to the rapid pace of development, it’s best to build Julia from source. To be able to keep up with the literally dozen code changes per day, you can clone the Julia GitHub repository to your local machine. If you use one of the GitHub GUI’s, this is as easy as hitting the ‘Sync Branch’ button to receive all of the newest code updates.

To install Julia, you need to compile the code. The instructions for each supported operating system are listed on the Julia GitHub page. For Mac users, use Terminal to navigate to the directory where you cloned Julia, then run the following command, where ‘n’ refers to the number of concurrent processes you want the compiler to use:

1
make -j n 

I use 8 concurrent processes on a 2013 MacBook Pro and it works pretty well. Certainly much faster than a single process. Note that the first time you run the make command, the build process will take much longer than successive builds, as Julia downloads all the required libraries needed. After the first build, you can just run the make command with a single process, as the code updates don’t take very long to build.

Package management is also done via GitHub. To add Julia packages to your install, you use the Pkg.add() function, with the package name in double-quotes.

Julia code feels very familiar

Text file import

Although the Julia documentation makes numerous references to MATLAB in terms of code similarity, Julia feels very familiar to me as an R and Python user. Take reading a .csv file into a dataframe and finding the dimensions of the resulting object

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
#R: Read in 1987.csv from airline dataset into a dataframe
#No import statement needed to create a dataframe in R
airline1987 <- read.csv("~/airline/1987.csv")
dim(airline1987)
[1] 1311826      29

#Python: use pandas to create a dataframe
import pandas as pd
airline1987 = pd.read_csv("/Users/randyzwitch/airline/1987.csv")
airline1987.shape
Out[7]: (1311826, 29)

#Julia: use DataFrames to create a dataframe
using DataFrames
airline1987 = readtable("/Users/randyzwitch/airline/1987.csv")
size(airline1987)
(1311826,29)

In each language, the basic syntax is to call a ‘read’ function, specify the .csv filename, then the defaults of the function read in a basic file. I also could’ve specified other keyword arguments, but for purposes of this example I kept it simple.

Looping

Looping in Julia is similar to other languages. Python requires proper spacing for each level of a loop, with a colon for each evaluated expression. And although you generally don’t use many loops in R, to do so requires using parenthesis and brackets.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
#Python looping to create a term-frequency dictionary

from collections import Counter

term_freq = Counter()
for word in english_dictionary:
  for url in url_list:
    if word in url_list:
      term_freq[word] += 1

#Julia looping to create a term-frequency dictionary

term_freq=Dict{String, Int64}()
for word in english_dictionary
    for url in url_list
        if search(line, word) != (0:-1)
            term_freq[word]=get(term_freq,word,0)+1
        end
    end
end

If you’re coming from a Python background, you can see that there’s not a ton of difference between Python looping into a dictionary vs. Julia. The biggest differences are the use of the end control-flow word and that Julia doesn’t currently have the convenience “Counter” object type. R doesn’t natively have a dictionary type, but you can add a similar concept using the hash package.

Vectorization

While not required to achieve high performance, Julia also provides the functional programming construct of vectorization and list comprehensions. In R, you use the *apply family of functions instead of loops in order to apply a function to multiple elements in a list. In Python, there are the map and reduce functions, but there is also the concept of list comprehensions. In Julia, both of the aforementioned functionalities are possible.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
#Cube every number from 1 to 100

#Python map function
cubes = map(lambda(x): x*x*x, range(1,100))

#Python list comprehension
cubes= [x*x*x for x in range(1,100)]

#R sapply function
cubes <- sapply(seq(1,100), function(x) x*x*x)

#Julia map function
cubes = map((x)-> x*x*x, [1:100])

#Julia list comprehension
cubes = [x*x*x for x in [1:100]]

In each case, the syntax is just about the same to apply a function across a list/array of numbers.

A small, but intense community

One thing that’s important to note about Julia at this stage is that it’s very early. If you’re going to be messing around with Julia, there’s going to be a lot of alone-time experimenting and reading the Julia documentation. There are also several other resources including a Julia-Users Google group, Julia for R programmers, individual discussions on GitHub in the ‘Issues’ section of each Julia package, and a few tutorials floating around (here and here).

Beyond just the written examples though, I’ve found that the budding Julia community is very helpful and willing in terms of answering questions. I’ve been bugging the hell out of John Myles White and he hasn’t complained (yet!), and even when code issues are raised through the users group or on GitHub, ultimately everyone has been very respectful and eager to help. So don’t be intimidated by the fact that Julia has a very MIT and Ph.D-ness to it…jump right in and migrate some of your favorite code over from other languages.

While I haven’t moved to using Julia for my everyday workload, I am getting facility to the point where I’m starting to consider using Julia for selected projects. Once the language matures a bit more, JuliaStudio starts to approach RStudio in terms of functionality, and I get more familiar with the language in general, I can see Julia taking over for at least one if not all of my scientific programming languages.

  • RSiteCatalyst Version 1.4.16 Release Notes
  • Using RSiteCatalyst With Microsoft PowerBI Desktop
  • RSiteCatalyst Version 1.4.14 Release Notes
  • RSiteCatalyst Version 1.4.13 Release Notes
  • RSiteCatalyst Version 1.4.12 (and 1.4.11) Release Notes
  • Self-Service Adobe Analytics Data Feeds!
  • RSiteCatalyst Version 1.4.10 Release Notes
  • WordPress to Jekyll: A 30x Speedup
  • Bulk Downloading Adobe Analytics Data
  • Adobe Analytics Clickstream Data Feed: Calculations and Outlier Analysis
  • Adobe: Give Credit. You DID NOT Write RSiteCatalyst.
  • RSiteCatalyst Version 1.4.8 Release Notes
  • Adobe Analytics Clickstream Data Feed: Loading To Relational Database
  • Calling RSiteCatalyst From Python
  • RSiteCatalyst Version 1.4.7 (and 1.4.6.) Release Notes
  • RSiteCatalyst Version 1.4.5 Release Notes
  • Getting Started: Adobe Analytics Clickstream Data Feed
  • RSiteCatalyst Version 1.4.4 Release Notes
  • RSiteCatalyst Version 1.4.3 Release Notes
  • RSiteCatalyst Version 1.4.2 Release Notes
  • Destroy Your Data Using Excel With This One Weird Trick!
  • RSiteCatalyst Version 1.4.1 Release Notes
  • Visualizing Website Pathing With Sankey Charts
  • Visualizing Website Structure With Network Graphs
  • RSiteCatalyst Version 1.4 Release Notes
  • Maybe I Don't Really Know R After All
  • Building JSON in R: Three Methods
  • Real-time Reporting with the Adobe Analytics API
  • RSiteCatalyst Version 1.3 Release Notes
  • Adobe Analytics Implementation Documentation in 60 Seconds
  • RSiteCatalyst Version 1.2 Release Notes
  • Clustering Search Keywords Using K-Means Clustering
  • RSiteCatalyst Version 1.1 Release Notes
  • Anomaly Detection Using The Adobe Analytics API
  • (not provided): Using R and the Google Analytics API
  • My Top 20 Least Useful Omniture Reports
  • For Maximum User Understanding, Customize the SiteCatalyst Menu
  • Effect Of Modified Bounce Rate In Google Analytics
  • Adobe Discover 3: First Impressions
  • Using Omniture SiteCatalyst Target Report To Calculate YOY growth
  • ODSC webinar: End-to-End Data Science Without Leaving the GPU
  • PyData NYC 2018: End-to-End Data Science Without Leaving the GPU
  • Data Science Without Leaving the GPU
  • Getting Started With OmniSci, Part 2: Electricity Dataset
  • Getting Started With OmniSci, Part 1: Docker Install and Loading Data
  • Parallelizing Distance Calculations Using A GPU With CUDAnative.jl
  • Building a Data Science Workstation (2017)
  • JuliaCon 2015: Everyday Analytics and Visualization (video)
  • Vega.jl, Rebooted
  • Sessionizing Log Data Using data.table [Follow-up #2]
  • Sessionizing Log Data Using dplyr [Follow-up]
  • Sessionizing Log Data Using SQL
  • Review: Data Science at the Command Line
  • Introducing Twitter.jl
  • Code Refactoring Using Metaprogramming
  • Evaluating BreakoutDetection
  • Creating A Stacked Bar Chart in Seaborn
  • Visualizing Analytics Languages With VennEuler.jl
  • String Interpolation for Fun and Profit
  • Using Julia As A "Glue" Language
  • Five Hard-Won Lessons Using Hive
  • Using SQL Workbench with Apache Hive
  • Getting Started With Hadoop, Final: Analysis Using Hive & Pig
  • Quickly Create Dummy Variables in a Data Frame
  • Using Amazon EC2 with IPython Notebook
  • Adding Line Numbers in IPython/Jupyter Notebooks
  • Fun With Just-In-Time Compiling: Julia, Python, R and pqR
  • Getting Started Using Hadoop, Part 4: Creating Tables With Hive
  • Tabular Data I/O in Julia
  • Hadoop Streaming with Amazon Elastic MapReduce, Python and mrjob
  • A Beginner's Look at Julia
  • Getting Started Using Hadoop, Part 3: Loading Data
  • Innovation Will Never Be At The Push Of A Button
  • Getting Started Using Hadoop, Part 2: Building a Cluster
  • Getting Started Using Hadoop, Part 1: Intro
  • Instructions for Installing & Using R on Amazon EC2
  • Video: SQL Queries in R using sqldf
  • Video: Overlay Histogram in R (Normal, Density, Another Series)
  • Video: R, RStudio, Rcmdr & rattle
  • Getting Started Using R, Part 2: Rcmdr
  • Getting Started Using R, Part 1: RStudio
  • Learning R Has Really Made Me Appreciate SAS