corARMA: ARMA(p,q) Correlation Structure

In pwr4exp: Power Analysis for Research Experiments

ARMA(p,q) Correlation Structure

Description

Re-exports nlme::corARMA from the nlme package.

Usage

corARMA(value, form, p, q, fixed)

Arguments

value	numeric vector of parameter values
form	a one-sided formula of the form ~ t, ~ 1 | g, or ~ t | g, specifying a time covariate t, a grouping factor g, or both. When no time covariate is specified, the row order of the data within each group is assumed to represent the chronological order of measurements.
p	non-negative integer specifying the autoregressive order
q	non-negative integer specifying the moving average order
fixed	unused

Details

In the original nlme::corARMA function, a covariate t must be an integer class. In pwr4exp, t can also be a factor class, which is then converted to an integer internally for chronological order. The class of t in the model formula, if present, is not converted. For example, a time covariate can be fitted as a factor in the model formula, whereas it is converted to an integer in the correlation formula.

See Also

corClasses See nlme::corARMA for original documentation.

pwr4exp documentation built on Nov. 11, 2025, 5:13 p.m.