[ENH] dunders for time series distances and kernels

[fkiraly]

This PR adds dunders for time series distances and kernels (descendants of BasePairwiseTransformerPanel), behaving as per likely user expectation, description as below.

It also adds tests for the different combinations of dunders and estimators.

algebraic operations between time series distances and kernels

• d = dist1 * dist2 satisfies d(X1, X2) == dist1(X1, X2) * dist2(X1, X2), for all pairwise transformers dist1, dist2, equality for all elements of the resulting matrix
• d = dist1 + dist2 gives d(X1, X2) == dist1(X1, X2) + dist2(X1, X2), for all pairwise transformers dist1, dist2
• for a pairwise distance dist and an int or float const, d = dist * const or d = const * dist gives d(X1, X2) == dist(X1, X2) * const
• for a pairwise distance dist and an int or float const, d = dist + const or d = const + dist gives d(X1, X2) == dist(X1, X2) + const

pipeline concatenation between ordinary transformers and time series distances or kernels

• for a transformer trafo and a pairwise dist, the esteimator pipe = trafo * dist is also a pairwise distance, with pipe(X1, X2) == dist(trafo.fit_transform(X1), trafo.fit_transform(X2))
• above, the transformer trafo can be an sktime transformer, or an sklearn transformer (which is coerced/wrapped)

This especially may be interesting for users with a research interest in time series classification or clustering, as it allows to obtain common time series distances easily as composition of others.

E.g., ddtw (for common definitions of ddtw) is the same as Differencer() * DtwDist() (first difference, then dtw distance). Higher order differences or other combinations are also easy to obtain by this.