estimate inverse probability weights based on Beesley & Mukhherjee

Usage

ipw(
  stacked_data,
  weight_outcome_var = "WTFA_A",
  samp_var = "samp_WTFA_A",
  external_dataset = "NHIS",
  dataset_name = "MGI",
  id_var = "id",
  cancer_factor = FALSE,
  cancer_factor_var = "cancer",
  covs = c("age_50", "female", "nhw", "hypertension", "diabetes", "cancer", "anxiety",
    "depression", "bmi_cat"),
  chop = TRUE
)

Arguments

stacked_data: data.table of stacked data
weight_outcome_var: variable name of sampling weights in external dataset
samp_var: variable name of sampling weights in internal dataset
external_dataset: name of external dataset
dataset_name: name of internal dataset
id_var: variable name of id
cancer_factor: logical, whether to include cancer factor
cancer_factor_var: variable name of cancer factor
covs: vector of covariates to include in model
chop: logical, whether to chop weights

Value

return table of id and inverse probability weights