rbmi (version 1.4.1)

set_vars: Set key variables

Description

This function is used to define the names of key variables within the data.frame's that are provided as input arguments to draws() and ancova().

Usage

set_vars(
  subjid = "subjid",
  visit = "visit",
  outcome = "outcome",
  group = "group",
  covariates = character(0),
  strata = group,
  strategy = "strategy"
)

Arguments

subjid

The name of the "Subject ID" variable. A length 1 character vector.

visit

The name of the "Visit" variable. A length 1 character vector.

outcome

The name of the "Outcome" variable. A length 1 character vector.

group

The name of the "Group" variable. A length 1 character vector.

covariates

The name of any covariates to be used in the context of modeling. See details.

strata

The name of the any stratification variable to be used in the context of bootstrap sampling. See details.

strategy

The name of the "strategy" variable. A length 1 character vector.

Details

In both draws() and ancova() the covariates argument can be specified to indicate which variables should be included in the imputation and analysis models respectively. If you wish to include interaction terms these need to be manually specified i.e. covariates = c("group*visit", "age*sex"). Please note that the use of the I() function to inhibit the interpretation/conversion of objects is not supported.

Currently strata is only used by draws() in combination with method_condmean(type = "bootstrap") and method_approxbayes() in order to allow for the specification of stratified bootstrap sampling. By default strata is set equal to the value of group as it is assumed most users will want to preserve the group size between samples. See draws() for more details.

Likewise, currently the strategy argument is only used by draws() to specify the name of the strategy variable within the data_ice data.frame. See draws() for more details.

See Also

draws()

ancova()

Examples

if (FALSE) {

# Using CDISC variable names as an example
set_vars(
    subjid = "usubjid",
    visit = "avisit",
    outcome = "aval",
    group = "arm",
    covariates = c("bwt", "bht", "arm * avisit"),
    strategy = "strat"
)

}