Higer dimesion unfolding

[davo417]

Hi, I'm a young researcher in particle physics and I've found this repo while looking for a machine learning approach to the unfolding problem.

The OmniFold method seems great and I've already played with the two demos provided here but I was wandering if, in order to do multi-dimensional unfolding, will me enough to concatenate multiple observables and then take slices from the unfolded data.

Another question I have is if I need the same amount of Monte Carlo (MC) data than real measured data in order to unfold a big dataset from an experiment. I could take batches from the data the same length as my MC samples and perform the unfolding on every batch, but that means I have to repeat step 1 and 2 every time? Is there something it can be reused from one batch to the next?