about sam about sewar HOT 4 CLOSED

Hello @jianyuan2015,

Would you please supply the codes you are comparing to?

from sewar.

In my opinion, if the size of input data GT is "[M, N, K]", where K is number of spectral band, the code shouble be like this:

GT = GT.reshape((GT.shape[0]*GT.shape[1],GT.shape[2]))
P = P.reshape((P.shape[0]*P.shape[1],P.shape[2]))
N = GT.shape[0]
sam_angles = np.zeros(N)
for i in range(GT.shape[0]):
      val = np.clip(np.dot(GT[i,:],P[i,:]) / (np.linalg.norm(GT[i,:])*np.linalg.norm(P[i,:])),-1,1)
      sam_angles[i] = np.arccos(val)

from sewar.

Hi @fkmajiji,

I think the current implementation is the correct one. We use the K (spectral band) as N. See example Matlab implementation here.

Closing this now. Feel free to reopen if you disagree

from sewar.

I'm sorry, I saw the example Matlab implementation you mentioned and still think the current implementation is not correct. The introduction of Matlab function in https://github.com/davidkun/HyperSpectralToolbox/tree/master/functions has said "N The total number of pixels." and "p Number of bands."

from sewar.