I managed to execute your demo notebook and the associated building_pointcloud_main.py script. However, there are a number of open questions I would like to speak with you about:
Neither the notebook nor the .py script complete successfully on the vm, see screenshot below. Have you been able to run it successfully on the vm before?
Based on the computed statistics, only 1.23% of the footprints in York have an associated pointcloud. Do you know why the ratio is so low?
The computed statistics say that 1264 pointclouds have been generated for York. However, in the npy_raw folder for York only 903 pointclouds are listed. This seems to be a bug.
I realize we are saving the .npy outputs inside the VM as of now. Let's save the outputs directly to the data_share folder, please. That way, we don't have to spin up the VM in order to access the generated outputs.
Let's address these issues on a new feature branch, please.