make slicing of Sequence more efficient

[DaniBodor]

I noticed that when slicing a Sequence, a we first create a deepcopy of the original object, which we then whittle down to the size indicated. This seems inefficient, especially if the original Sequence is much larger than the slice we want.

Instead we could create a new "empty" Sequence, into which we then copy (or deepcopy?) the relevant slices from the old Sequence.

Does this make sense @psomhorst ?

[psomhorst]

Yes, this makes a lot of sense.

I'm wondering what the best approach is. I defaulted to deepcopy to make sure the Framesets, Events, etc. were copied, instead of referencing to the same Framesets, Events, etc. as the original. Maybe we can create an implementation of Sequence.__copy__() (see bottom paragraph of https://docs.python.org/3.11/library/copy.html#copy.deepcopy) to create a copy of Sequence that (deep)copies all Events, Phases, etc., but creates new Framesets using sliced data instead of copying the entire Frameset including data.

[DaniBodor]

Also, see this discussion on ensuring that copy.deepcopy(sequence) and sequence.deepcopy() are equivalent.