Anglemapper

From Eigenvector Research Documentation Wiki
Revision as of 16:19, 10 September 2020 by Scott (talk | contribs) (Created page with "===Purpose=== Classification based on angle measures between signals. ===Synopsis=== :[x,angl] = anglemapper(x,targ,options) ===Description=== Anglemappercan be used with...")
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

Purpose

Classification based on angle measures between signals.

Synopsis

[x,angl] = anglemapper(x,targ,options)

Description

Anglemappercan be used with two-way matrices (e.g., time-series) or multivarite images from imaging spectroscopy. Samples are classed as the target with the smallest angle between the sample and the target. Different algorithms can be selected using (options.algorithm). The default is a simple angle measure (a.k.a., Spectral Angle Mapper).

For a two-way MxN matrix and a set of K target signals or spectra (targ), ANGLEMAPPER calculates an angle between each target and each row of the input (x). The output (x) is a MxN DataSet object with the last class for Mode 1 (x.class{1,end}) corresponding to the target with the smallest angle.

For an MxN DataSet object (x) of type "image" with a total of M pixels and a set of K target spectra (targ), ANGLEMAPPER calculates an angle between each target and each pixel (row) the input (x). The output (x) is a MxN image DataSet object with the last class for Mode 1 (x.class{1,end}) corresponding to the target with the smallest angle.

Inputs

  • x = data set (MxN class double or DataSet object). Images must be input as a DataSet object (see buildimage).
  • targ = targets (KxN class double or DataSet object) with K targets, K>1. If (targ) is a DataSet object, the output new class will inherit the classname and class lookup table from the desired target class using (options.classset).

Optional Inputs

  • display : [ 'off' | {'on'} ] governs level of display and waitbar.
  • algorithm : [ {sam} | 'scm' | 'wtfa' | 'wtfac' ]
  • threshold : [{80}], threshold in degrees for classification [0<threshold<90]. Samples/Pixels with all target angles >threshold are not given a class.
* sam :  Spectral Angle Mapper (SAM): Yuhas,RH, Goetz, AFH, Boardman, JW, "Discrimination Among Semi-Arid Landscape Endmembers Using Spectral Angle Mapper (SAM) Algorithm," Summaries of the 4th Annual JPL AirborneG eoscience Workshop, JPL Pub-92-14, AVIRIS Workshop. Jet Propulsion Laboratory, Pasadena, CA, 147-150 (1992).
* scm :  Spectral Correlation Mapper (SCM): de Carvalho Jr, OA, Meneses, PR, "Spectral Correlation Mapper (SCM): An Improvement on the Spectral Angle Mapper (SAM)," Summaries of the 9th JPL Airborne Earth Science Workshop, JPL, Publication 00-18, 9 p. (2000).
* wtfa :  Window Target FActor Analysis (wtfa): Lohnes, MT, Guy, RD, Wentzell, PD, "Window Target-Testing Factor Analysis: Theory and Application to the Chromatographic Analysis of Complex Mixtures with Multiwavelength Fluorescence Detection", Anal. Chim. Acta, 389, 95-113 (1999). Malinowski, ER, "Obtaining the Key Set of Typical Vectors by Factor Analysis and Subsequent Isolation of Component Spectra," 134, 129- 137 (1982). DOI: 10.1016/S0003-2670(01)84184-2 Malinowski, ER, Factor Analysis in Chemistry, 2nd ed. John Wiley & Sons, New York 1991.
* wtfac =  Window Target FActor Analysis with Correlation (wtfac).

Outputs

  • x = input (x) with a new class added to the sample mode as a MxN DataSet object.
  • angl = MxK matrix of angles for each of the K targets.

See Also

buildimage, evolvfa, ewfa, ewfa_img, pca, wtfa, wtfa_img