Returns a tibble containing parameter estimates from a bbi_{.model_type}_summary object. Details about the tibble that is returned are in the return value section below. Note: if you need to pull in the parameter estimates for a large number of NONMEM models at once, consider using param_estimates_batch() instead.

param_estimates(.summary, .alpha)

# S3 method for bbi_nonmem_summary
param_estimates(.summary, .alpha = NULL)

Arguments

.summary

A bbi_{.model_type}_summary object.

.alpha

numeric. If specified, return the p-value for each eta value on the diagonal, as well as whether they are below the specified value of alpha. Defaults to NULL.

Value

Returns a tibble with the following columns:

  • parameter_names -- Parameter name ("THETA1", "THETA2", etc.)

  • estimate -- Parameter estimate

  • stderr -- Standard Error

  • random_effect_sd -- OMEGA and SIGMA elements in standard deviation/correlation format

  • random_effect_sdse -- Standard errors to the OMEGA and SIGMA elements in standard deviation/correlation format

  • fixed -- TRUE if parameter is fixed, FALSE otherwise

  • diag -- TRUE if parameter is a diagonal element in random effect matrix, FALSE if off-diagonal, NA if parameter is a THETA

  • shrinkage -- Shrinkage for final estimation method, using SD parameterization (for NONMEM this is ETAshrinkSD and EPSshrinkSD). The shrinkage using Variance Parameterization, as well as for other estimation methods and multiple subpops if present, can be found in bbi_nonmem_summary$shrinkage_details.

Details

Note that Bayesian methods are not yet supported by this function. Creating a parameter table like this for Bayesian estimation methods requires a different approach, which has not yet been implemented. If the final estimation method of the input model is Bayesian, a not implemented error will be thrown.

Methods (by class)