Describe the feature you'd like to see

For some large countries, users may be willing to perform a clustering in higher details for a region/district, rather than the entire country (i.e. India).
The gadm data contains in flags "NAME_1" or "NAME_2" the names of region and/or the district that could be used to automatically identify the regions of interest.

Those namings could be used to tailor the focus_weights of the model, see #789