Fixes for calc_shape_dist()

[astamm]

I found a couple of issues in find_rotation_seed_coord() that had repercussions on calc_shape_dist(). The latter now performs the same calculations as in curve_pair_align(). In particular, the object beta2best output by find_rotation_seed_coord() is properly rotated and scaled when rotation == TRUE and scale == TRUE which makes it directly useable in calc_shape_dist() and curve_pair_align() without the need to perform the rotation, scaling and group action again a posteriori.

calc_shape_dist() is also better documented and its output has been augmented.

• I have not renamed already existing output components in order not to break existing code that uses the function, although I think that d and dx for amplitude and phase distances could have better names.
• I added betascale which is intended to be the optimal scaling factor that has been applied to the second curve when scale == TRUE; it is nothing but the length of beta1 divided by the length of beta2;
• we could output qscale as well but it is not currently the case.

Lastly, I feel that we should document that the output beta1 is slightly different from the input of the same name. by clarifying what the centring performed via calculatecentroid() does.