nmmds

Non-metric multi-dimensional scaling - Kruskal's Non-Metric MDS

Input Arguments:

• DISSIM is a vector of dissimilarities, containing the upper triangular part of an interpoint distance matrix, similar to what one would get from PDIST.
• D is the dimensionality of the lower-dimensional space.
• R is the Minkowski metric. A value of R=1 gives the city-block distance, and R=2 yields the Euclidean.

Output Arguments:

• X is the observations in the d-dimensional space.
• STRESS is the value of the stress.
• DHAT is the disparaties.

Synopsis

[X, STRESS,DHAT] = NMMDS(DISSIM,D,R)

References:

Cox, Trevor F. and Michael A. A. Cox. 2001. Multidimensional Scaling, 2nd Edition, Boca Raton: Chapman & Hall/CRC.

Kruskal, Joseph B. 1964a. “Multidimensional scaling by optimizing goodness of fit to a nonmetric hypothesis,” Psychometrika, 29:1-27.

Kruskal, Joseph B. 1964b. “Nonmetric multidimensional scaling: A numerical method,” Psychometrika, 29:115-129.