Correct unsupervised version

[sylvchev]

When no labels are provided, it is possible to use the unsupervised loss. The implemented version is not correct:

riemannianPCA/riemannianpca.py

Line 142 in 277e4b9

y = np.array([-1] * (n_samples // 2) + [1] * (n_samples // 2))

Intermediate possibility is to use only one label for all the data.