Abundance maps functions¶
FCLS¶
-
pysptools.abundance_maps.amaps.
FCLS
(M, U)¶ Performs fully constrained least squares of each pixel in M using the endmember signatures of U. Fully constrained least squares is least squares with the abundance sum-to-one constraint (ASC) and the abundance nonnegative constraint (ANC).
- Parameters:
- M: numpy array
- 2D data matrix (N x p).
- U: numpy array
- 2D matrix of endmembers (q x p).
- Returns: numpy array
- An abundance maps (N x q).
- References:
- Daniel Heinz, Chein-I Chang, and Mark L.G. Fully Constrained Least-Squares Based Linear Unmixing. Althouse. IEEE. 1999.
- Notes:
- Three sources have been useful to build the algorithm:
- The function hyperFclsMatlab, part of the Matlab Hyperspectral Toolbox of Isaac Gerg.
- The Matlab (tm) help on lsqlin.
- And the Python implementation of lsqlin by Valera Vishnevskiy, click: http://maggotroot.blogspot.ca/2013/11/constrained-linear-least-squares-in.html , it’s great code.
NNLS¶
-
pysptools.abundance_maps.amaps.
NNLS
(M, U)¶ NNLS performs non-negative constrained least squares of each pixel in M using the endmember signatures of U. Non-negative constrained least squares with the abundance nonnegative constraint (ANC). Utilizes the method of Bro.
- Parameters:
- M: numpy array
- 2D data matrix (N x p).
- U: numpy array
- 2D matrix of endmembers (q x p).
- Returns: numpy array
- An abundance maps (N x q).
- References:
- Bro R., de Jong S., Journal of Chemometrics, 1997, 11, 393-401.