G.start {cds}

Constrained Dual Scaling for a Single Random G Start

Description

Run algorithm for a single G matrix.

Usage

G.start(X, nr.starts.a, astarts, maxit, n, m, q, Fr.cent, maxit.ALS, Mmat,
  eps.G, info.level, times.a.multistart, eps.ALS, const, K, random.G, tol,
  update.G)

Arguments

X

List of two elements, namely i giving the number of the start and G given the starting configuration

nr.starts.a

The number or random starts for a to use in the ALS.

astarts

Explicit starts for a, if applicable.

maxit

The maximum number of iterations with respect to G.

n

The number of respondents.

m

The number of items.

q

The maximum rating category such that the rating scale is 1:q.

Fr.cent

The centred Fr matrix.

maxit.ALS

The maximum number of ALS iterations.

Mmat

The basis matrix for the quadratic monotone splines.

eps.G

The absolute error tolerance for the G updates.

info.level

Integer controlling the amount of information printed.

times.a.multistart

The number of times random starts for a is used.

eps.ALS

The absolute error tolerance for the ALS.

const

The constant part of the loss function.

K

The number of groups.

random.G

The random argument passed to updateG.

tol

tolerance tol passed to lsei of the limSolve package)

update.G

Logical indicating whether or not to update the starting configuration G in X