home *** CD-ROM | disk | FTP | other *** search
- function [fcal, f] = pcrcal(a, c, vc, n)
- % PCRCAL Perform P-matrix type quant by cross-correlation
- %
- % PCRCAL does the P-matrix type quant by cross-correlation
- % in the chosen abstract vector space.
- %
- % fcal = pcrcal(a, c, vc, n) or [fcal, f] = fcalf(a, c, vc, n)
- %
- % Where:
- %
- % fcal is the calibration matrix
- % f is the matrix containing the calibration in factor space
- % a is the absorbance matrix
- % c is the concentration matrix
- % vc is the matrix containing the factors (eigenvectors)
- % n is the number of factors to use in the calibration
- %
- % The concentrations of components in unknown spectra can be
- % calculated if the unknown absorbances are set up as a vector (for 1
- % unknown) or a matrix (for multiple unknowns). The unknown matrix
- % can have any available name. For example call it "u". The
- % unknown concentrations are then given by fcal * u.
- %
-
- % Copyright (c) 1989-92 by The MathWorks, Inc.
-
- [t, maxrank] = size(vc);
- if n > maxrank, n = maxrank, end;
- vc = vc(:, 1:n);
- proj = (vc' * a);
- f = c * proj' * inv(proj * proj');
- fcal = f * vc';
-