concorR (version 0.2.1)

concor1: Find +1/-1 convergence points of an input matrix

Description

A subfunction of concor.

Runs the Pearson correlation on the matrix stack repeatedly until a square matrix consisting of only +1 and -1 is formed.

Usage

concor1(m_stack, cutoff = .9999999, max_iter = 50)

Arguments

m_stack

The stack of input matrices, which must include all transposes and relations.

cutoff

The absolute value of the convergence threshold for each matrix element.

max_iter

Maximum number of times to run the correlation while seeking convergence.

Value

A square matrix, with number of rows/columns equal to the number of columns of m_stack, where the values are either +1 or -1. The input matrix can later be split into two "positions" based off the locations of the positive/negative values.

Details

For network data with R relations, the input matrix "stack" is a (2 x N x R) x N matrix, consisting of each relation's adjacency matrix, then the transpose of that matrix, appended to each other vertically. The correlation is run until either all matrix entries have absolute values greater than cutoff or the maximum number of iterations max_iter is reached. On its own, this function does not execute the whole CONCOR method, but is listed separately from concor to demonstrate the core step.

See Also

concor

Examples

# NOT RUN {
a <- matrix(c(0, 0, 0, 1, 1, 0, 1, 0, 1, 1, 0, 1, 1, 0, 1, 0), ncol = 4)
b <- rbind(a, t(a))

concor1(b)
# }