add sample_id to distributed file splitting

[oHunewald]

Adding a sample_id column to the splitted files or in a separated vector in each worker to re-construct the relation between the training data after splitting.

[laurentheirendt]

@oHunewald, still relevant?

[exaexa]

I guess this is still relevant although not that pressing. A standalone sample_id vector can be constructed manually using a tiny modification of the current data loading code. Guess we could just add a helper function?