Update Python Version?

[ed-p-may]

Hi,

I am trying to use this image but having some trouble:

I need to use at least Python3.10 for my project, but since the Image here uses Ubuntu 20.04, it seem to only have Python3.8 installed when running this Container? I have been unable to figure out how to install an updated Python on top of this base Image for my own Containers. I have tried many methods, but none seem to work for me for some reason.

Therefor I wonder if it is it possible to update this Image to use the latest Ubuntu version (22.04 I think?) which would then update the installed Python version to 3.10?

I cannot build the Image myself with an updated OS, as I am just on a Mac computer.

Python 3.8 is nearing end of life status, so it seems like it would make sense to include an upgrade to a modern supported version somehow? I don't know enough about Images to figure out how to upgrade it myself, so I'm wondering if that can be included in your base image somehow?

thanks!
@ed-p-may

[bonnema]

Would the latest release help? https://github.com/NREL/EnergyPlus/releases/tag/v24.1.0

[ed-p-may]

Hi @bonnema ,

Thanks. Yes I'd love to use 24.1.0 - is there an updated Image for it available? I wasn't sure? The latest Image appeared to be EnergPlus 23.2 ?

[bonnema]

Let me ask around, it's probably as simple as someone updating a version string. Stay tuned!

[ed-p-may]

Thanks @bonnema !

I was trying it the other way around too (install EnergPlus on an existing Python 3.10+ Image ) but I can't figure out how to get past the 'Accept License enter 'y'' in the docker build.... If there is a way to update these Images though, that would be really helpful!

[Myoldmopar]

I can't figure out how to get past the 'Accept License enter 'y'' in the docker build.

FWIW, on that aspect I would consider not using the EnergyPlus "install" where you have to interactively run the EnergyPlus shell script installer, but instead just download the tar.gz and extract it. It's got the same stuff in it.

[Myoldmopar]

Therefor I wonder if it is it possible to update this Image to use the latest Ubuntu version (22.04 I think?) which would then update the installed Python version to 3.10?

I am sure the Docker image could be updated to Ubuntu 22.04 -- or even 24.04 in a few weeks maybe 😉 . But if you need a newer Python, it's quite easy to use pyenv to manage multiple Python versions. You could easily run the pyenv install script, use pyenv install 3.10 or newer, and be able to run it happily.

[Myoldmopar]

Would the latest release help?

I don't think the latest release of E+ will make a difference for this specific issue. While we do build against Python (3.8 for now), and include the Pyhon lib inside the guts of the EnergyPlus install, I think this is more about the system level Python.

But yes I'm sure there can be a new Docker image set up for E+ 24.1 soon!

[ed-p-may]

thanks @Myoldmopar !

Yes you are exactly right: I was referring to the system level python installation specification - that's the one my other script code is using.

My hope was just to get a Container that had a) a working install of E+, and also b) a Python3.10+ install that my scripts can utilize.

When using FROM nrel/energyplus:23.2.0 as the Image base, it gave me only Python-3.8. And since it uses Ubuntu-20, it appears there is no Python-3.10 package available to update it with. But when using FROM ubuntu:22.04 as the Image Base, while that worked on the Python side, it then gave me a lot of trouble installing E+....

---

FWIW, I tried the 'tar.gz' extract method, and while that seemed to work when building the Image: I ran into other errors related to missing 'usr/lib64/' files when I tried to actually use E+ in the Container. I'm not familiar with that install method - do you happen to know if I am supposed to setup other path variables or run any other 'setup' scripts if I just unzip the tar directly like that?

I like your idea about using pyenv to just install the new Python on top of the existing Ubuntu-20 install - I had been try to use the 'deadsnake' installation to do the same thing but was unable to make it work. I'm not familiar with using PyEnv in a Dockerfile though? Are you aware of any examples that show its usage?

@ed-p-may