Open aloboa opened 3 months ago
The result of endmembers depends on your sample and the settings/choices of MNF and PPI. For rock samples, this looks reasonable. However, for more explorative work, I would recommend being less restrictive in the MNF correlation or eigenvalues reduction and on the PPI definition. That way, you will have more than one endmember (but potentially more potential noise). We will try to describe this process in the upcoming documentation. At this point, we can discuss this via PM.
Both PPI and Compute end-members end with
AttributeError: 'SpectralPlotter' object has no attribute 'figure'
In the case of the endmembers, is this the plot that I am supposed to get?