Skip to contents

Weight diagnostics for IPTW

Source: R/weightR.R
weightR.Rd

Compute weights and diagnostic plots and mean differences

Usage

weightR(
  data,
  model,
  treatment,
  vars,
  labs.headings = list(),
  labs.subheadings = list(),
  difference = "standardized",
  weights = "w",
  tail = 50,
  num.vars,
  simplify = F,
  digits = 2
)

Arguments

data

data.frame

model

logistic regression model for estimating propensity scores

treatment

treatment variable

vars

vector of covariates to estimate weights

labs.headings

list of varible names in the format list(new = "old")

labs.subheadings

list of variable levels in the format list(var = list(new = "old"))

difference

whether the results should be outputtet as absolute or standardized weighted differences

weights

name of the column containing weights

tail

the number of covariate-groups with the highest weights to show (e.g. top 50)

num.vars

pseudo numerical variables for ordering of levels

simplify

drop first level of binary variables

Value

list containing a table of standardized differences (table), overall propensity scores and weights (overall), balance plot (balance) and weight plots (weightst)