Create Propensity Score Model and Associated Plots — margot_propensity_model_and

This function creates a propensity score model using the specified exposure variable and baseline covariates. It also generates associated plots and diagnostics.

Usage

margot_propensity_model_and_plots(
  df_propensity,
  exposure_variable,
  baseline_vars,
  weights_var_name,
  estimand = "ATE",
  method = "ebal",
  focal = NULL,
  love_plot_options = list(),
  bal_tab_options = list(),
  verbose = TRUE
)

Arguments

df_propensity: A data frame containing the variables for the propensity score model.
exposure_variable: A character string specifying the name of the exposure variable in df_propensity.
baseline_vars: A character vector specifying the names of the baseline covariates in df_propensity.
weights_var_name: A character string specifying the name of the weights variable in df_propensity.
estimand: A character string specifying the estimand. Default is "ATE" (Average Treatment Effect).
method: A character string specifying the method for propensity score estimation. Default is "ebal".
focal: For binary treatments, a value of the treatment to be considered "focal" (i.e., the intervention). Default is NULL.
love_plot_options: A list of options to be passed to cobalt::love.plot(). Default is an empty list.
bal_tab_options: A list of options to be passed to cobalt::bal.tab(). Default is an empty list.
verbose: A logical value indicating whether to print progress messages. Default is TRUE.

Value

A list containing the following elements:

match_propensity: The propensity score model object.
love_plot: A love plot object created by cobalt::love.plot().
summary: A summary of the propensity score model.
summary_plot: A plot of the propensity score model summary.
balance_table: A balance table created by cobalt::bal.tab().
diagnostics: A list of additional diagnostic information.

Examples

if (FALSE) { # \dontrun{
# Assuming df_propensity is your dataset with appropriate variables
results <- margot_propensity_model_and_plots(
  df_propensity = df_propensity,
  exposure_variable = "treatment",
  baseline_vars = c("age", "sex", "bmi"),
  weights_var_name = "sample_weights",
  love_plot_options = list(
    thresholds = c(m = .05),
    size = 4
  ),
  bal_tab_options = list(
    thresholds = c(m = .1, v = 2.5)
  ),
  verbose = TRUE
)

# Access the results
print(results$summary)
plot(results$love_plot)
print(results$balance_table)
} # }