JuliaPrecompile

Code to precompile mostly used packages for data science. Uses PackageCompiler.jl to make a sys image, which helps to deal with the time to first plot issue.

You would probably want to make your own version with a personalised package list.

Usage

First clone this repository:

git clone https://github.com/bartosz-zbik/JuliaPrecompile.git

Then compile wanted packages by running:

julia scripts/compile_jds.jl

This may take a few minutes.

Finally add an alias or a symlink in your bin pointing to run_jds.sh

Versions and Package List

Currently, there is only one version available:

Data Science (jds)

Packages that will be compiled:

• DataFrames
• CSV
• StatsPlots
• GLM
• Statistics (stdlib)

Packages that will load automatically at startup:

• Statistics (stdlib)
• DataFrames
• StatsBase
• GLM
• StatsPlots
• LinearAlgebra (stdlib)
• CSV (but only the module, you need to do CSV.read())
• Revise (only if present, and only in interactive mode)

No matter which version you choose the environment also contains:

• StatsKit (and it components)
• Latexify
• LaTeXStrings
• IJulia for Jupyter integration
• some SciML packages

For the full list of packages see Project.toml file.