selectable clustering variable for a dataset; sample data

[vineetbansal]

• Translation issue with aligned volume should be fixed now (offset all slices using translations for the very first layer, which means that the original volume and aligned volume will match w.r.t. the positions of the top layer).
• The adata.obs variable that defines clustering ("original_clusters" in the sample data, but could be anything going forward), is now selectable from the UI.
• Sample data included with the plugin (uploaded to Dropbox, downloaded through the pooch library)
• Opening local files as slices through open files as stack.. brings up the plugin UI automatically.
• unit tests for napari plugin
• Support for one or multiple overlap_fraction values (one for each adjacent pair of slices).
• Moved all dataset-related classes to paste3.dataset. Added some (basic) tests.

[vineetbansal]

@anushka255 - this is ready from my end. I'll send another PR with minor tweaks by the end of this week.