Display failure

[Dharmesh946]

Hi

After successfully running the training and the following stage, the script fails in viz_scrna.py, line 460 with

AttributeError: module 'matplotlib.cm' has no attribute 'get_cmap'.

Probably a matplotlib version issue (mine is 3.9.0).

[nekonekko]

Hi

matplotlib.cm.get_cmap() has been removed in matplotlib==3.9.0. (See: API Changes for 3.9.0 — Matplotlib 3.9.0 documentation)

You can use matplotlib.pyplot.get_cmap() instead of matplotlib.cm.get_cmap().

[Dharmesh946]

Thanks, I implemented the fix.
Now I'm getting a lot of warnings

viz_scrna.p                                                                                                                     y:458: UserWarning: The input coordinates to pcolormesh are interpreted as cell centers,                                                                                                                      but are not monotonically increasing or decreasing. This may lead to incorrectly calculat                                                                                                                     ed cell edges, in which case, please supply explicit cell edges to pcolormesh.
  plt.pcolormesh(xx, yy, qz)

Do you think its normal? I could not go further yet to a missing dependency on my side (ffmpeg) (btw, I think it would be interesting to improve the readme with the requirements).

In order to move on my tests, without needing to run again the training each time, how can I run the last part of the code (displaying the results) on the generated data?