Encountered code bugs while running the Jupyter Notebook demos in the tutorial

[Daiqy]

Hi @aviadlevis ,

Thank you for your excellent work!

I have configured the environment according to the installation steps in the README, and then tried to run the jupyter notebooks in the tutorials. However, I encountered the following bugs:

• Tutorial2 - synthesize ngEHT observations.ipynb

  • Code:
  • Error:

• tutorials/Tutorial3 - estimate 3D emission from image plane.ipynb

  • Code:
  • Error:

Additionally, here are the versions of the packages installed in my conda virtual environment (Python 3.9.19, CUDA 12.0, Ubuntu 20.04): packages.txt

I have tried some solutions, including adjusting the versions of some packages, but still haven't been able to successfully run the programs. Could you please offer some suggestions? Thanks for your attention to this issue!

Best regards, Daiqy

[aviadlevis]

Thanks a lot Daiqy!

The first error seems to be coming form eht-imaging. Could you pase the outputs of running obs_empy.tlist()? What is the data structure?

The second error is easy to fix, I think that this .unfreeze() is deprecated so you can just return params. I will fix this in the github soon.

[Daiqy]

Thanks for your quick reply!

The first error: I ran the obs_empty.tlist() before obs = bhnerf.observation.observe_same(movie, obs_empty), and got the output as followed:

The second error: I removed .unfreeze() and successfully ran the iteration; however, I've encountered the new error (perhaps still a version issue?):