medflex (version 0.6-10)

plot.neModel: Confidence interval plots for natural effect components

Description

Obtain effect decomposition confidence interval plots for natural effect models.

Usage

# S3 method for neModel
plot(x, xRef, covLev, level = 0.95, transf = identity, ylabels, yticks.at, ...)
# S3 method for neModelBoot
plot(
  x,
  xRef,
  covLev,
  level = 0.95,
  ci.type = "norm",
  transf = identity,
  ylabels,
  yticks.at,
  ...
)

Arguments

x

a fitted natural effect model object.

xRef

a vector including reference levels for the exposure, x* and x, at which natural effect components need to be evaluated (see details).

covLev

a vector including covariate levels at which natural effect components need to be evaluated (see details).

level

the confidence level required.

transf

transformation function to be applied internally on the (linear hypothesis) estimates and their confidence intervals (e.g. exp for logit or Poisson regression). The default is identity (i.e. no transformation).

ylabels

character vector containing the labels for the (linear hypothesis) estimates to be plotted on the y-axis.

yticks.at

numeric vector containing the y-coordinates (from 0 to 1) to draw the tick marks for the different estimates and their corresponding confidence intervals.

...

additional arguments.

ci.type

the type of bootstrap intervals required (see type argument in neModel-methods).

Details

This function yields confidence interval plots for the natural effect components. These causal parameter estimates are first internally extracted from the neModel object by applying the effect decomposition function neEffdecomp(x, xRef, covLev).

Examples

data(UPBdata)

impData <- neImpute(UPB ~ att * negaff + educ + gender + age, 
                    family = binomial, data = UPBdata)
neMod <- neModel(UPB ~ att0 * att1 + educ + gender + age, 
                 family = binomial, expData = impData, se = "robust")

plot(neMod)
plot(neMod, transf = exp, 
     ylabels = c("PDE", "TDE", "PIE", "TIE", "TE"))
plot(neMod, level = 0.9, xRef = c(-1, 0))