Open APJansen opened 6 months ago
This is done in 3689512
It has two remaining issues:
For the first point, it may be easiest to change the way the train/val split is done. Currently a selection is made on samples and days for the trail+val set, and then randomized along the heights. Perhaps it's better to do the train/val split on days instead.
This needs to be rewritten to be compatible with 3D Unet. It should just take the yaml file with the coordinates and dynamically extract the data in the patch size relevant for the model at hand.