I am completely new to julia, and I would like to learn how to optimize my code to run as fast as possible. I don’t know if this has already been answered or not, but I created a .julia/config/startup.jl file with the line
@time using DifferentialEquations, Plots;
so that everytime I open the repl I can just use the package functions. Everytime I open the repl it takes 30 seconds to load. Is there a way to use a precompiled version to make it faster?
3 Likes
Check out:
2 Likes
How does this relate to https://github.com/JuliaLang/PackageCompiler.jl?
What is my best option if I’m interested in distributable shared libaries? ( .dll/.so )
Please excuse the necropost, but this question didn’t get an answer and I want to set the record straight (although someone more knowledgeable might supply details) in this thread
You should use GitHub - JuliaLang/PackageCompiler.jl: Compile your Julia Package .
The version with “X” appended is an obsolete developement version.
The closest you can get is also PackageCompiler.jl.
3 Likes
No worries. Always happy to have an answer. I have some issues at julia github that are going on a year old, would be happy for some necro-updates to that 
The old PackageCompiler.jl used to support building shared library files with c exported functions of my choice. This new direction seems very focused on the time-to-first-plot problem and not too concerned with my use case. I had trouble getting it to export my own Base.@ccallable function which wasn’t a problem with the older one.
I’ll raise issues there if I can’t tweak it to produce what I want.
Thanks.