Wtfa img

From Eigenvector Research Documentation Wiki
Jump to navigation Jump to search

Purpose

Window target factor analysis for images.

Synopsis

[rho,angl,q] = wtfa_img(spec,tspec,window,p,options)

Description

Inputs

  • spec = MxNxP 3-way data matrix/image (spatial modes 1 and 2 MxN, and spectral mode is 3 w/ P channels).
  • tspec = KxP matrix of target/candidate spectra.
  • window = 2 element vector containing the window width in the x- and y-directions {each element should be > 1} (Note: if a scalar is input then the window in both directions is set to the scalar).
p = The number of principal components, PCs, for modeling each window of spectra.
p >= 1: (integer) number PCs is a constant p,
0 < p < 1: sets a relative criterion for selecting number of PCs in each window i.e. only the first set of PCs that together capture >=p*100% of the variance in the window are used
p < 0: sets an absolute value for number of PCs i.e. factors with singular values <|p| are not used.

Outputs

  • rho = direction cosine between tspec and a p component PCA model of (spec) in each window.
  • angl = angle between targets and PCA model [= acos(rho)].
  • q = Q residuals.

Options

options = a structure array with the following fields:

  • plots: [ 'off' | {'angle'} | 'rho' | 'q' ]
'off' : nothing is plotted,
'angle' : projection angle (angle) plotted {default},
'rho' : direction cosine (rho) are plotted, or
'q' : Q residuals (q) are plotted.
  • inds: [ ] a 2 column matrix of indices corresponding to the location in the image to test {default uses entire image}.

See Also

evolvfa, ewfa, ewfa_img, pca, wtfa