dot-plot_balance: (For maintainer) Plot balance

In trangnguyen74/tnMediation: Simple and more robust estimation of marginal natural (in)direct effects

(For maintainer) Plot balance

Description

Internal functions that makes balance plots for weighted data.

Usage

.plot_balance.med(w.dat, c.vars, m.vars, c.std, m.std, key.balance = FALSE)

.plot_balance.ipw(w.dat, vars, vars.std)

.plot_balance.odds(w.dat, vars, vars.std)

.plot_balance.psYpred(w.dat, c.vars, c.std)

.plot_balance.psYpredMR(w.dat, cross.world, c.vars, m.vars, c.std, m.std)

.plot_balance.YpredMR(w.dat, cross.world, c.vars, m.vars, c.std, m.std)

.plot_balance.wpMRCadj(w.dat, c.vars, m.vars, c.std, m.std)

Arguments

w.dat	Data for (pseudo) subsamples, e.g. output of .compute_weights.Ypred()
c.vars	Names of covariates, already cleaned.
m.vars	Names of mediators, already cleaned.
c.std	Names of covariates whose mean differences are to be standardized, already cleaned.
m.std	Names of mediators whose mean differences are to be standardized, already cleaned.
key.balance	Whether all balance components are crucial to the estimator using the weighting method, defaults to FALSE. See relevance in Value section.
vars	Names of variables whose balance is to be plotted, already cleaned.
vars.std	Names of variables whose mean differences are to be standardized, already cleaned.
cross.world	(For plot_balance.psYpredMR) blah

Value

Plot of balance on the means of covariates and mediators between pseudo subsample(s) and relevant subsample(s).

Plot of balance on the means of covariates and mediators between relevant pseudo samples and full sample. If estimate_wtd==TRUE, add "(anchor)" and "(for <effect>)" notes to plot labels to draw attention to how each balance matters to the estimator.

trangnguyen74/tnMediation documentation built on May 3, 2023, 6:58 a.m.