powerStressMin {stops}

Power Stress SMACOF

Description

An implementation to minimize power stress by minimization-majorization. Usually more accurate but slower than powerStressFast.

Usage

powerStressMin(
  delta,
  kappa = 1,
  lambda = 1,
  nu = 1,
  weightmat = 1 - diag(nrow(delta)),
  init = NULL,
  ndim = 2,
  acc = 1e-10,
  itmax = 50000,
  verbose = FALSE
)

Arguments

delta

dist object or a symmetric, numeric data.frame or matrix of distances

kappa

power of the transformation of the fitted distances; defaults to 1

lambda

the power of the transformation of the proximities; defaults to 1

nu

the power of the transformation for weightmat; defaults to 1

weightmat

a matrix of finite weights

init

starting configuration

ndim

dimension of the configuration; defaults to 2

acc

numeric accuracy of the iteration

itmax

maximum number of iterations. Defaults to 50000.

verbose

should iteration output be printed; if > 1 then yes

Value

an object of class 'smacofP' (inheriting form 'smacofB', see smacofSym). It is a list with the components

• delta: Observed dissimilarities, not normalized

• obsdiss: Observed transformed dissimilarities, not normalized

• confdist: Configuration dissimilarities, NOT normalized

• conf: Matrix of fitted configuration, NOT normalized

• stress: Default stress (stress 1; sqrt of explicitly normalized stress)

• spp: Stress per point (based on stress.en)

• ndim: Number of dimensions

• model: Name of smacof model

• niter: Number of iterations

• nobj: Number of objects

• type: Type of MDS model

• weightmat: weighting matrix

• pars: hyperparameter vector theta

and some additional components

• stress.m: default stress is the explicitly normalized stress on the normalized, transformed dissimilarities

• deltaorig: observed, untransformed dissimilarities

• kappa: kappa parameter

• lambda: lambda parameter

• nu: nu parameter (aka rho)

Note

The functionality related to power stress and the 'smacofP' class is also available in the 'cops' package. Expect masking when both are loaded.

Author(s)

Jan de Leeuw & Thomas Rusch

See Also

smacofSym

Examples

dis<-smacof::kinshipdelta
res<-powerStressMin(as.matrix(dis),kappa=2,lambda=1.5,nu=2,
                    weightmat=as.matrix(dis/2),itmax=1000)
res
summary(res)
plot(res)