uniscale: Unidimensional Scaling from Seriation Results

Description

Performs (approximate) unidimensional scaling by first performing seriation to obtain a permutation and the using the permutation to calculate the configuration.

Usage

uniscale(d, order = NULL, method = "QAP_LS", rep = 10, ...)
orderplot(x, main, pch = 19, ...)

Value

A vector with the fitted configuration.

Arguments

d: a dissimilarity matrix.
order: a precomputed permutation (configuration) order. If NULL, then seriation is performed using the method specified in method.
method: seriation method used if o is NULL.
rep: Number of repetitions of the seriation heuristic.
...: additional arguments are passed on to the seriation method.
x: a scaling returned by uniscale().
main: main plot label
pch: print character

Author

Michael Hahsler with code from Patrick Mair (from smacof).

Details

Uses the method describes in Maier and De Leeuw (2015) to calculate the minimum stress configuration for either a given configuration/permutation/order or for a permutation computed via a seriation method.

The code is similar to uniscale() in smacof, but scales to larger datasets since it does not check all permutations.

References

Mair P., De Leeuw J. (2015). Unidimensional scaling. In Wiley StatsRef: Statistics Reference Online, Wiley, New York. tools:::Rd_expr_doi("10.1002/9781118445112.stat06462.pub2")

Examples

Run this code

data(SupremeCourt)

d <- as.dist(SupremeCourt)

sc <- uniscale(d)
sc

orderplot(sc)

Run the code above in your browser using DataLab