I’d say yes, in your cluster.
I don’t recommend changing the environment variable for the rest of your system, I’d create an alias for Julia, or a small shell wrapper
I’d say yes, in your cluster.
I don’t recommend changing the environment variable for the rest of your system, I’d create an alias for Julia, or a small shell wrapper
I guess you are right.
actually, most of my use is to run julia scripts. So adding LD_LIBRARY_PATH=""
as a prefix before running the script is quite easy.
so apparently running
LD_LIBRARY_PATH="" julia script.jl
does not make the errors go away.
Namely, I still see this error in the output (since I am importing PyPlot):
┌ Warning: No working GUI backend found for matplotlib
└ @ PyPlot ~/.julia/packages/PyPlot/XHEG0/src/init.jl:165
@giordano, do you know what should be done in this case?
Is there a way to set it from “inside” the script?
To be clear, you are experiencing multiple different unrelated issues:
LD_LIBRARY_PATH
environment variable being set which causes JLL packages to not be loaded correctlyMy suggestion would be to use some other plotting libraries. Honestly I’m not familiar with internals of PyPlot.jl
to know what’s going on. If you really want to use PyPlot.jl
I suggest you to open a new thread, this has been resolved long ago (the original problem with the git tree sha1 was solved by upgrading to Julia v1.6)