Validate a pmx_stats container

Checks that x is a well-formed pmx_stats container and returns it invisibly, or aborts with an error of class pmxhelpr_error_bad_stats naming what is wrong. Use it after editing a container (e.g. x$stats) or building one by hand with pmx_stats(); is_pmx_stats() with strict = TRUE runs the same checks but returns TRUE/FALSE.

The generic dispatches on the container class:

• vpc_stats: the bin, count and quantile columns of stats; the OBSDV/PC_OBSDV/TIME columns of obs; the n_replicates, ci, pi, loq and strat_var config keys; and the sim_prop_blq_* columns when config$loq is set.

• doseprop_stats: the power-model columns of stats, an obs frame holding the config metric and dose columns, and the metric_name_var, metric_value_var, dose_var and ci config keys.

• forest_stats: the estimate, label, is_ref and layout columns of stats with their types and values, the config grouping columns, and valid config$ci and config$sigdigits.

• Any other object: x must carry the pmx_stats class and have a stats data.frame, an obs data.frame or NULL, and a config list. Packages defining their own pmx_stats subclass can add a method.

Usage

# S3 method for class 'doseprop_stats'
validate_pmx_stats(
  x,
  ...,
  arg = rlang::caller_arg(x),
  call = rlang::caller_env()
)

# S3 method for class 'forest_stats'
validate_pmx_stats(
  x,
  ...,
  arg = rlang::caller_arg(x),
  call = rlang::caller_env()
)

validate_pmx_stats(
  x,
  ...,
  arg = rlang::caller_arg(x),
  call = rlang::caller_env()
)

# Default S3 method
validate_pmx_stats(
  x,
  ...,
  arg = rlang::caller_arg(x),
  call = rlang::caller_env()
)

# S3 method for class 'vpc_stats'
validate_pmx_stats(
  x,
  ...,
  arg = rlang::caller_arg(x),
  call = rlang::caller_env()
)

Arguments

x

Object to validate.

...

Passed to methods; unused by the methods in pmxhelpr.

arg, call

The argument name and call reported in errors.

Value

invisible(x) on success.

See also

Other pmx stats class: is_pmx_stats(), pmx_stats()

Examples

fs <- df_forest(data_sad_pkforest)
validate_pmx_stats(fs)
fs$stats$est <- as.character(fs$stats$est)
try(validate_pmx_stats(fs))
#> Error in eval(expr, envir) : 
#>   `fs$stats` column `est` must be numeric.