pspaf_discrete: Internal, pathway specific PAF when the mediator is discrete

In graphPAF: Estimating and Displaying Population Attributable Fractions

Internal, pathway specific PAF when the mediator is discrete

Description

Internal, pathway specific PAF when the mediator is discrete

Usage

pspaf_discrete(
  data,
  refval,
  riskfactor_col,
  mediator_col,
  mediator_model,
  response_model,
  weight_vec
)

Arguments

data	dataframe. A dataframe (with no missing values) containing the data used to fit the mediator and response models. You can run data_clean to the input dataset if the data has missing values as a pre-processing step
refval	For factor valued risk factors, the reference level of the risk factor. If the risk factor is numeric, the reference level is assumed to be 0
riskfactor_col	Integer indicator for the risk factor column in data
mediator_col	Integer indicator for the discrete mediator column in data
mediator_model	A glm or polr model for the mediator, depending on the same confounders and risk factor as specified in the response model.
response_model	A R model object for a binary outcome that involves a risk factor, confounders and mediators of the risk factor outcome relationship. Note that a weighted model should be used for case control data. Non-linear effects should be specified via ns(x, df=y), where ns is the natural spline function from the splines library.
weight_vec	A numeric column of weights

Value

A numeric vector (if ci=FALSE), or data frame (if CI=TRUE) containing estimated PS-PAF for each mediator referred to in mediator_models, together with estimated direct PS-PAF and possibly confidence intervals.

graphPAF documentation built on May 29, 2024, 10:21 a.m.