Concatenate processed datasets

This MR adds a new optional argument in concat_dataset and multi_experiment_data_extractor to allow the concatenation of multiple processed datasets.

Currently, these functions allow us to concatenate raw datasets within a certain range of TUIDs. These data directories my also contain analysis results folders which contain processed datasets. We may wish to concatenate the processed datasets as well as the raw datasets.

I have added a new optional analysis_name argument to both of these functions. If analysis_name is None, we simply concatenate the raw datasets. If an analysis name is specified, we take the processed dataset from the analysis folder with the specified name, and we concatenate these datasets. There is also a new dimension optional argument in multi_experiment_data_extractor, because the processed datasets may have to be concatenate along dimensions with names other than dim_0.

Blocked by !393 (merged)