gtrunsec / bat-julia-tutorial

BAT/Julia tutorial for DESY Statistics School

BAT/Julia Tutorial at Advanced School on Statistics in HEP, DESY, Oct. 2019

Welcome to the BAT/Julia tutorial.

Software installation

Before beginning this tutorial, please install ensure you have Jupyter, Julia and BAT.jl installed and everything is set up correctly.

Running on your laptop

Follow the installation instructions in the BAT.jl documentation.

You will find installers for Jupyter, Julia, source-code editors and other software on your tutorial USB pen-drive.

Running on CERN SWAN

Using a few tricks, it's possibly to install Julia on CERN SWAN and run BAT.jl there.

Note: For the purposes of this tutorial, installing and running Julia on your laptop will likely be more convenient and responsive.

Warning: The following procedure seems to work, but has not been extensively tested - use at your own risk!

Start a SWAN environment with software stack "96 Python3". Then source the script cern-swan-init-julia.sh within a SWAN terminal session (source ./cern-swan-init-julia.sh). This should install Julia into "$HOME/sw/julia" and register the IJulia kernel with the SWAN Jupyter instance.

Afterwards, you should be able to run the command julia in the terminal session and install all Julia packages required for this tutorial:

julia> using Pkg
julia> pkg"add BAT ArraysOfArrays Distributions ElasticArrays IntervalSets Parameters Plots ValueShapes StatsBase Tables TypedTables HDF5" 

See the BAT.jl documentation for details.

Reload the SWAN "My Projects" web-page before starting Julia notebooks.

Running the tutorial

Please clone the tutorial Git repository

git clone https://github.com/bat/bat-julia-tutorial.git

(If you're using SWAN, ensure that "bat-julia-tutorial" ends up in the "SWAN_projects" directory. It's easiest to use the "Download Project from git" function on the "My Projects" web-page).

Start a jupyter server using either

jupyter lab

or

jupyter notebook

depending on which Jupyter UI style you prefer.

Start the notebook "Exercise-1.ipynb" to begin the tutorial.

Learning Julia

Due to the format of the tutorial, you should (hopefully) not need to consult documentation frequently. Still, here are a few resources that should help you get started with the Julia language quickly:

• The "MATLAB–Python–Julia cheatsheet" show MATLAB, Python and Julia code for common operations in direct comparison.

• "The Fast Track to Julia" is a convenient online cheatsheet for quick reference.

• The official Julia language documentation.

• "Think Julia" is a good, detailed online book about the Julia language.

• The Julia homepage has a large section with introductory material.