PCA

[AndreaGarciaJuan]

When using profiles with few number of points: z = np.arange(0.,-100,-10.), pyxpcm.fit rise an error during PCA calculations

[gmaze]

Can you give more information about the PCA error ?

I suspect this is because with z = np.arange(0.,-100,-10.), you only have 10 points, and thus, reducing the dimensions is not really necessary, So the PCA fails because it may ask for 12 PCA, while they are only 10 levels

So: increase the nb of points, or avoid dimensionality reduction, using the reduction=0 when you create the pcm.

[AndreaGarciaJuan]

ValueError: n_components=15 must be between 0 and min(n_samples, n_features)=10 with svd_solver='full' This is the error

[gmaze]

this is exactly my answer, not enough levels, useless reduction