get_sigma function

Extract residual variability parameter estimates from a NONMEM output object.

get_sigma(x, output = "est", sigdig = 6, sep = "-", est.step = NULL)

Arguments

• x: A NONMEM output object generated using read_nm.
• output: A flag specifying the matrix or matrices to be output. Valid flag values are est (the default), se, rse, cor, cse, 95ci, or all.
• sigdig: Specifies the number of significant digits to be provided (default=6).
• sep: Specifies the separator character to use for 95% confidence intervals (default="-").
• est.step: Specifies which estimation step to return parameters from (default is the last).

Returns

A symmetrical matrix, or a list of symmetrical matrices if all is specified.

est returns the estimated SIGMA variance-covariance matrix. se returns the standard errors for the estimated SIGMA variance-covariance matrix. rse returns the relative standard errors for the estimated SIGMA variance-covariance matrix (se/est100). cor returns the correlation matrix matrix. cse returns the standard errors for the correlation matrix. 95ci returns the asymptotic 95% confidence intervals for the elements of the SIGMA variance-covariance matrix (est +/- 1.96se). all returns all available SIGMA matrices.

Examples

## Not run:

 nmOutput <- read_nm("run315.xml")
 sigmas   <- get_sigma(nmOutput)
 sigmaRSEs <- get_sigma(nmOutput, "rse")
## End(Not run)

See Also

NONMEM (https://www.iconplc.com/innovation/nonmem/)

Author(s)

Justin Wilkins, justin.wilkins@occams.com