Closed aminravanbakhsh closed 2 years ago
Hi @aminravanbakhsh, This is normal behaviour. It is updating packages, not updating Julia itself.
You can disable this behaviour with update=False
.
Also note that this only occurs the first time you run fit(X, y)
in a Python process. Subsequent calls will not do this. If you quit and re-open Python, it will happen again.
Best, Miles
Hi Miles, Thanks for your response. However, I used "update = False" and it didn't work. Do you have any other ideas about this?
Regards, Amin
Okay, that seems like a mistake. I just looked and indeed Pkg.instantiate()
is called even when update=False
.
I pushed a fix to master.
Can you try with the version on master? i.e., pip install -U git+https://github.com/MilesCranmer/PySR.git@master
.
Note that since this is a new version, the first time you run it, you need to have update=True
, so that the new environment can be installed. Subsequent calls can have update=False
, though.
Thanks for your attention. Unfortunately it didn't work for me. Do you have any other ideas?
Can you try now? (i.e., re-run the pip install
?)
Note that it will still display the Activating project
message. It shouldn't display the Updating registry
message now though.
Unfortunately it didn't work too. I tried to reinstall the PySR. I received this warning, maybe it could help.
@aminravanbakhsh can you be more specific about what’s not working? From my test it seems like everything is good to go.
I am just saying that every time I run the code, it updates the registry, even when I reinstalled the PySR. I am just using a virtual environment on my Ubuntu 20.04.
It might not have updated your pysr
since it is the same version before and after the fix. Can you run pip uninstall pysr && pip install -U git+https://github.com/MilesCranmer/PySR.git@master
?
You are using update=False
, right?
Thanks for your helps Miles! It worked and I didn't receive the update message. However, it took a lot of time to start the process. Is it OK? Moreover, I am interested in how PySR works. Is there any paper that describes the PySR algorithm?
Awesome - yeah the first time you run it will take a while since Julia is compiling the backend. The second time you call .fit()
(within the same Python instance) will be much faster though.
Paper out soon!
Hi, I hope you are doing well. I used PySR to solve the example on the PySR github page. However, each time I run the code, it upgrades Julia. I would appreciate it if you could tell me what the problem is. It is worth mentioning that I use a virtual environment for my code. You can see the results below.
Version (please include the following information):