Interested in using Julia for Computer Vision. I've been playing with this in Python, but I've had a sinking feeling I would evntually run into performance issues. :P
Herein I stumble through creating and structuring a project, managing dependencies, and all of the other goodies. This is a "kitchen sink" repo where I install bunch of Julia libraries and play with them individually.
I'm on Mac OS X.
This README.md is notes and links from scattered Internet posts and documentation for my future reference.
curl -fsSL https://install.julialang.org | sh
julia --project="." src/hello.jl
or just:
make hello
Nice post about how to create and structure a Julia project.
mkdir juliacv/
cd juliacv
mkdir src/
Opens the REPL:
julia --project="."
Alias:
make shell
In the REPL:
using Pkg
Pkg.add("CSV")
Pkg.add("DataFrames")
Pkg.add("Plots")
Pkg.add("LinearAlgebra")
You can also install a package by trying to import it, and following the "should I just install that?" prompt that comes up.
using LinearAlgebra
Confirm install by entering package manager mode (PMM):
]
And then typing:
st
There has to be a better way of doing this, but in PMM mode:
compat
In CMM, I believe you do this to install all the listed dependencies:
activate .
instantiate
In PMM mode:
up
and update each package with the major and minor result of st. Also set the version of Julia you are using.
Jupyter Notebook kernel:
Pkg.add("IJulia")
and just install jupyter with pip.
Getting started with Julia and OpenCV
Adding OpenCV bindings:
Pkg.add("OpenCV")
Pkg.add("CxxWrap")
Flux for Deeping Learning in Julia
using Flux
freeCodeCamp post on doing ML with Julia.
using Metalhead