I\'m implementing the sklearn CCA classifier but I think my setup is flawed. For training, the input has 63 features and 81 samples (depending on the training m
sklearn