Estimate variable importance and counter-factual plots of a glm.nb model

Usage

inv_variables_glm_nb(
  cv_set,
  y_var,
  id_vars,
  pred_vars,
  var_scales,
  constant_org = NULL,
  constant_date = NULL,
  nsim = 50
)

Arguments

cv_set: list object containing analysis and assessment data.frames
y_var: string name of column of observed values
id_vars: names of columns used as keys/ids in data.frames
pred_vars: vector string of predictor variables to use in model
var_scales: data.frame containing centering and scaling parameters for variables
constant_org: which orgUnit to use for counterfactual plots. If NULL (default), chooses one at random
constant_date: whcih date to use for counterfactual plots. If NULL (default), is set to median of analysis data.
nsim: number of simulations to use in variable importance permutation. Default = 50.

Value

list containing variable importance scores and a list of dataframes containing data for counterfactual plots with each element corresponding to a variable