[ENH] Added useful class attributes to SAST based transforms

[IRKnyazev]

Reference Issues/PRs

Fixes #2004

What does this implement/fix? Explain your changes.

Added class attributes found in ST that are helpful for interpreting shapelets. They do not serve functionality for the classifier but improve explainability.

Implemented: The class each subseries is from, the index of the original series the subseries is extracted from, the start pos of the subseries in the original series.