Dims in PyMC models are just labels for when we convert draws to InferenceData. They don’t have any effect on the model (broadcasting and indexing still follow numpy logic and are positional based). This is because PyTensor doesn’t work like Xarray
Dims in PyMC models are just labels for when we convert draws to InferenceData. They don’t have any effect on the model (broadcasting and indexing still follow numpy logic and are positional based). This is because PyTensor doesn’t work like Xarray