reshape_vec2mat: Assemble a variance-covariance matrix
In psychmeta: Psychometric Meta-Analysis Toolkit

reshape_vec2mat

R Documentation

Assemble a variance-covariance matrix

Description

The reshape_vec2mat function facilitates the creation of square correlation/covariance matrices from scalars or vectors of variances/covariances. It allows the user to supply a vector of covariances that make up the lower triangle of a matrix, determines the order of the matrix necessary to hold those covariances, and constructs a matrix accordingly.

Usage

reshape_vec2mat(
  cov = NULL,
  var = NULL,
  order = NULL,
  var_names = NULL,
  by_row = FALSE,
  diag = FALSE
)

Arguments

`cov`	Scalar or vector of covariance information to include the lower-triangle positions of the matrix (default value is zero). If a vector, the elements must be provided in the order associated with concatenated column (`by_row = FALSE; default`) or row (`by_row = TRUE`) vectors of the lower triangle of the desired matrix. If variances are included in these values, set the `diag` argument to `TRUE`.
`var`	Scalar or vector of variance information to include the diagonal positions of the matrix (default value is 1).
`order`	If cov and var are scalars, this argument determines the number of variables to create in the output matrix.
`var_names`	Optional vector of variable names.
`by_row`	Logical scalar indicating whether `cov` values should fill the lower triangle by row (`TRUE`) or by column (`FALSE`; default).
`diag`	Logical scalar indicating whether `cov` values include variances (`FALSE` by default; if `TRUE`, the variance values supplied with the `cov` argument will supersede the `var` argument).

Value

A variance-covariance matrix

Examples

## Specify the lower triangle covariances
## Can provide names for the variables
reshape_vec2mat(cov = c(.3, .2, .4), var_names = c("x", "y", "z"))

## Specify scalar values to repeat for the covariances and variances
reshape_vec2mat(cov = .3, var = 2, order = 3)

## Give a vector of variances to create a diagonal matrix
reshape_vec2mat(var = 1:5)

## Specify order only to create identity matrix
reshape_vec2mat(order = 3)

## Specify order and scalar variance to create a scalar matrix
reshape_vec2mat(var = 2, order = 3)

## A quick way to make a 2x2 matrix for bivariate correlations
reshape_vec2mat(cov = .2)

psychmeta documentation built on Aug. 26, 2022, 5:14 p.m.