Inline julia.magic kills Jupyter notebook (only on Linux) [pyjulia]

General Usage
,
1

I’ve noticed that Julia’s performance seems much better on my Linux partition than my Windows one (same GPU, Windows runs on my SSD while Linux is on and HDD). I was hoping to work on a project I have that’s partially written in Python and partially in Julia but loading inline Julia in my Jupyter notebook on Linux always kills the kernel.

import juila
%load_ext julia.magic

fails.

I got the error about statically linked Python and tried loading the API without compiled modules.

from julia.api import Julia
j = Julia(compiled_modeuls=False)

also fails.

Any hints?

2

Is that an actual partition, i.e. real Linux (as in distro) dual-booting? Or did you mean WSL2 (also “real” Linux [kernel], but not yet tier 1 support in Julia)? It seems not, but worth to try both.

I don’t know much about this, but maybe it applies for you to use juliacall from PythonCall.jl instead of pyjulia?

3

It’s a real Linux partition with partitioned disk space booted as “Fedora” from the UEFI screen. I’ll attempt with JuilaCall…