Most of the data variables are of shape (n_outputs, n_samples), but it would be nice to figure out if this is the correct assumption in all functions or if the data is transposed somewhere.
Concretely, this could involve adding a comment on top of each function that has vectors/matrices involved, what is:
- the assumed shape
- the true shape
Most of the data variables are of shape
(n_outputs, n_samples), but it would be nice to figure out if this is the correct assumption in all functions or if the data is transposed somewhere.Concretely, this could involve adding a comment on top of each function that has vectors/matrices involved, what is: