| author |
|---|
dimiboeckaerts |
Let's get going with Julia! This summary-based guide into programming with Julia is meant for students and researchers with a general background in programming but no specific knowledge of Julia. The first two sections get you up and running in Julia, while the third and fourth section dive into some more (bio)engineering-oriented material. Finally, the last section provides additional resources to explore more from Julia. Happy learning!
[TOC]
Let's start at the beginning. What is Julia again and why should you learn about it?
OK, now let's download & install Julia. You can either download the language itself and engage with it through the terminal using the built-in Julia command-line (or notebooks, see below), or you can download JuliaPro (requires a free account) which comes with Juno, an IDE to work in. A third way of using Julia is in Visual Studio Code, through a Julia extension.
First baby steps around the block: https://docs.julialang.org/en/v1/manual/getting-started/
Learn the basics of Julia: https://juliabyexample.helpmanual.io/ https://learnxinyminutes.com/docs/julia/
Learn more about types, multiple dispatch and structs (not needed to understand into detail for STMO): https://towardsdatascience.com/how-to-learn-julia-when-you-already-know-python-641ed02b3fa7
Coupling Julia to Jupyter notebooks is done with IJulia: https://www.youtube.com/watch?v=oyx8M1yoboY
Pluto is a lightweight, reactive notebook for Julia, with a lot of useful features: https://github.com/fonsp/Pluto.jl
During STMO, Jupyter notebooks will be used during the exercises, while Pluto will be used for the project.
Plots are really straightforward in Julia: https://docs.juliaplots.org/latest/tutorial/ http://ucidatascienceinitiative.github.io/IntroToJulia/Html/PlotsJL
Optimization in Julia is organized through projects in the JuMP and JuliaDiff communities, which coordinate the development of a wide breadth of functionality in mathematical programming, optimization and Operations Research. For more details, see the STMO course (https://github.com/MichielStock/STMO)!
Julia, JuMP and JuliaDiff packages address a variety of optimization problem categories (e.g. linear, nonlinear, semidefinite, second-order conic, and mixed-integer), provide functionality for advanced acceleration techniques (automatic differentiation and dual-number calculations), as well as provide high-level, embedded, domain-specific algebraic modeling languages directly in Julia (source: juliacomputing.com)
Some fun examples with JuMP: https://www.juliaopt.org/notebooks/JuMP-Sudoku.html https://www.juliaopt.org/notebooks/JuMP-Rocket.html
JuMP overview & hands-on tutorial: https://www.youtube.com/watch?v=7LNeR299q88 https://github.com/mfalt/juliacourse/blob/master/lecture6/JuMP_Overview.ipynb
A detailed overview of optimization in Julia in the book Julia Programming for Operations Research: https://www.softcover.io/read/7b8eb7d0/juliabook/frontmatter
Another family of packages for non-linear optimization in Julia is JuliaNLSolvers: https://github.com/JuliaNLSolvers
A good documentation on Julia tools for bioinformatics and computational biology is hard to find at first. BioJulia (the BioPython of Julia) seems like the only and all-encompassing Julia library for these kinds of tasks. https://github.com/BioJulia https://www.youtube.com/watch?v=6CpPd6tkokQ
- BioSequences.jl to work with DNA & protein sequences
- BioAlignments.jl to do alignments
- BioTools.jl to BLAST
- FASTX.jl to read & write FASTA & FASTQ files
- Phylogenies.jl to do phylogenetics
- ...
Another framework for some bioinformatics-related tasks in Julia is the MiToS package. It handles multiple sequence alignments, protein structures, alignments from Pfam and more. https://github.com/diegozea/MIToS.jl
Thirdly, ProteinEnsembles is a package that creates ensembles of protein structures: https://github.com/jgreener64/ProteinEnsembles.jl
Fourthly, simulate DNA sequencing experiments with Pseudoseq.jl: https://github.com/bioinfologics/Pseudoseq.jl
You can also do regular expressions in Julia: https://docs.julialang.org/en/v1/manual/strings/#Regular-Expressions also see: https://juliabyexample.helpmanual.io/#String-Manipulations
Here's an introductory video on general data analysis with Data Frames: https://www.youtube.com/watch?v=XRClA5YLiIc
Some interesting case studies:
- RNA sequencing data analysis: https://juliacomputing.com/case-studies/rna.html
- Medical diagnosis with AI: https://juliacomputing.com/case-studies/contextflow.html
- Precision medicine: https://juliacomputing.com/case-studies/pathbio.html
For starters, Scikit-learn has a Julia implementation! https://github.com/cstjean/ScikitLearn.jl
MLJ is another extensive framework for ML models in Julia that appears to comprise Scikit-learn: https://github.com/alan-turing-institute/MLJ.jl
Tutorials (on general data science & MLJ): https://alan-turing-institute.github.io/DataScienceTutorials.jl/ https://www.youtube.com/watch?v=ByFglWPqNlg
There are quite a few deep learning frameworks to work with in Julia: https://analyticsindiamag.com/top-9-machine-learning-frameworks-in-julia/
Flux and TensorFlow are two of the main DL frameworks. Knet is a third popular framework. https://juliacomputing.com/industries/ml-and-ai.html
An overview of how DL is fresh in Julia with tutorial: https://www.youtube.com/watch?v=4sOQN6cLqEc
Parallel & distributed computing in Julia: https://www.youtube.com/watch?v=JoRn4ryMclc https://github.com/mfalt/juliacourse/blob/master/lecture7/distributed.pdf
GPU computing in Julia: https://www.youtube.com/watch?v=7Yq1UyncDNc
Performance and profiling in Julia: https://github.com/mfalt/juliacourse/blob/master/lecture3/performance.pdf https://github.com/crstnbr/JuliaWorkshop19/blob/master/1_One/3_specialization.ipynb
During the STMO course, you can ask question on our Gitter: https://gitter.im/STMOUGent/community
A Julia (vs Python vs MATLAB) cheatsheet! https://cheatsheets.quantecon.org/
Free courses at Julia Academy: https://juliaacademy.com/courses
Community: https://julialang.org/community/