dot-point_est: (For maintainer) Point estimation

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

(For maintainer) Point estimation

Description

Internal functions to be called within estimate_\* to obtain the point estimate of effects.

Usage

.point_est.wtd(
  data,
  cross.world,
  effect.scale,
  a.c.form,
  a.cm.form,
  max.stabilized.wt = 30,
  output.data = FALSE
)

.point_est.psYpred(
  data,
  cross.world,
  effect.scale,
  a.c.form,
  max.stabilized.wt = 30,
  output.data = FALSE,
  y.c1.form,
  y.c0.form,
  y.cm1.form,
  y.cm0.form,
  y.family
)

.point_est.psYpredMR(
  data,
  cross.world,
  effect.scale,
  a.c.form,
  a.cm.form,
  max.stabilized.wt = 30,
  output.data = FALSE,
  y.c1.form,
  y.c0.form,
  y.cm1.form,
  y.cm0.form,
  y.family
)

.point_est.Ypred(
  data,
  cross.world,
  effect.scale,
  a.cm.form,
  max.stabilized.wt = 30,
  output.data = FALSE,
  y.c1.form,
  y.c0.form,
  y10.c.form = NULL,
  y01.c.form = NULL,
  y.family
)

.point_est.YpredMR(
  data,
  cross.world,
  effect.scale,
  a.c.form,
  a.cm.form,
  max.stabilized.wt = 30,
  output.data = FALSE,
  y.c1.form,
  y.c0.form,
  y10.c.form = NULL,
  y01.c.form = NULL,
  y.family
)

.point_est.Y2pred(
  data,
  cross.world,
  effect.scale,
  y.c1.form,
  y.c0.form,
  y.cm1.form,
  y.cm0.form,
  y10.c.form,
  y01.c.form,
  y.family,
  output.data = FALSE
)

.point_est.Y2predR(
  data,
  cross.world,
  effect.scale,
  a.c.form,
  a.cm.form,
  max.stabilized.wt = 30,
  output.data = FALSE,
  y.c1.form,
  y.c0.form,
  y.cm1.form,
  y.cm0.form,
  y10.c.form,
  y01.c.form,
  y.family
)

.point_est.NDEpred(
  data,
  cross.world,
  y.c1.form,
  y.c0.form,
  y.cm1.form,
  y.cm0.form,
  nde0.c.form,
  nde1.c.form,
  y.family,
  output.data = FALSE
)

.point_est.MsimYpred(
  data,
  cross.world,
  effect.scale,
  m.vars,
  m.c1.form,
  m.c0.form,
  m.family,
  y.c1.form,
  y.c0.form,
  y.cm1.form,
  y.cm0.form,
  y.family,
  output.data = FALSE,
  boot = FALSE,
  point.reps = 100
)

.point_est.wtCadj(
  data,
  cross.world,
  effect.scale,
  a.c.form,
  a.cm.form,
  max.stabilized.wt = 30,
  output.data = FALSE,
  wkng.form,
  y.family
)

.point_est.wpCadj(
  data,
  cross.world,
  effect.scale,
  a.c.form,
  max.stabilized.wt = 30,
  output.data = FALSE,
  y.cm1.form,
  y.cm0.form,
  wkng.form,
  y.family
)

.point_est.MsimYpredMR(
  data,
  cross.world,
  effect.scale,
  a.c.form,
  a.cm.form,
  max.stabilized.wt,
  m.vars,
  m.c1.form,
  m.c0.form,
  m.family,
  y.c1.form,
  y.c0.form,
  y.cm1.form,
  y.cm0.form,
  y.family,
  output.data = FALSE,
  boot = FALSE,
  point.reps = 100
)

.point_est.wpMRCadj(
  data,
  cross.world,
  effect.scale,
  a.c.form,
  a.cm.form,
  max.stabilized.wt = 30,
  output.data = FALSE,
  y.cm1.form,
  y.cm0.form,
  wkng.form,
  y.family
)

.point_est.NDEpredR(
  data,
  cross.world,
  a.c.form,
  a.cm.form,
  max.stabilized.wt = 30,
  output.data = FALSE,
  y.c1.form,
  y.c0.form,
  y.cm1.form,
  y.cm0.form,
  nde0.c.form,
  nde1.c.form,
  y.family
)

Arguments

data	A dataset that has been prepared to have the sampling with variable named s.wt (e.g., using function .clean_inputs_generic().
cross.world	Three options: "10", "01" or c("10", "01").
effect.scale	Three options: "MD", "MR", "OR".
a.c.form	Formula for the P(A|C) model (the propensity score model).
a.cm.form	Formula for the P(A|C,M) model.
max.stabilized.wt	Max stabilized weight allowed. Larger weights are truncated to this level. Default is 30.
output.data	Whether to output the weighted data in addition to the estimated potential outcome means and effects. Defaults to FALSE.
y.c1.form, y.c0.form	Model formulas for E[Y|C,A=1], E[Y|C,A=0], checked.
y.cm1.form, y.cm0.form	Model formulas for E[Y|C,M,A=1], E[Y|C,M,A=0], checked.
y.family	GLM (quasi-) family for outcome models.
y10.c.form, y01.c.form	Model formulas for E[Y(1,M0)|C] and E[Y(0,M1)|C].
wkng.form	Formula for working regression model

Value

If plot==FALSE, effect estimates (numeric vector). If plot==TRUE, a list including this vector of effect estimates and the weight distribution and mean balance plots.

trangnguyen74/mediationClarity documentation built on May 4, 2023, 6:22 a.m.