CausalMediation: CausalMediation

In AdrianAntico/RemixAutoML: AutoQuant

CausalMediation

Description

CausalMediation utilizes models from regmedint package

Usage

CausalMediation(
  data,
  OutcomeTargetVariable = NULL,
  TreatmentVariable = NULL,
  MediatorVariable = NULL,
  Covariates = NULL,
  MM_TreatmentCovariates = NULL,
  OM_TreatmentCovariates = NULL,
  OM_MediatorCovariates = NULL,
  SurvivalEventVariable = NULL,
  UnTreated_ReferenceIndicator = NULL,
  Treated_ReferenceIndicator = NULL,
  Mediator_ControlDirectEffectLevel = NULL,
  Covariate_NaturalDirectIndirect = 0,
  MediatorTargetType = "linear",
  OutcomeTargetType = "linear",
  TreatmentMediatorInteraction = TRUE,
  CaseControlSourceData = FALSE,
  RemoveNA = FALSE
)

Arguments

data	Data frame containing the following relevant variables.
OutcomeTargetVariable	yvar in underlying model. A character vector of length 1. Outcome variable name. It should be the time variable for the survival outcome.
TreatmentVariable	avar in underlying model. A character vector of length 1. Treatment variable name.
MediatorVariable	mvar in underlying model. A character vector of length 1. Mediator variable name.
Covariates	For main model
MM_TreatmentCovariates	emm_ac_mreg in underlying model. A character vector of length > 0. Effect modifiers names. The covariate vector in treatment-covariate product term in the mediator model.
OM_TreatmentCovariates	emm_ac_yreg in underlying model. A character vector of length > 0. Effect modifiers names. The covariate vector in treatment-covariate product term in the outcome model.
OM_MediatorCovariates	emm_mc_yreg in underlying model. A character vector of length > 0. Effect modifiers names. The covariate vector in mediator-covariate product term in outcome model.
SurvivalEventVariable	eventvar in underlying model. An character vector of length 1. Only required for survival outcome regression models. Note that the coding is 1 for event and 0 for censoring, following the R survival package convention.
UnTreated_ReferenceIndicator	a0 in underlying model. A numeric vector of length 1. The reference level of treatment variable that is considered "untreated" or "unexposed".
Treated_ReferenceIndicator	a1 in underlying model. A numeric vector of length 1.
Mediator_ControlDirectEffectLevel	m_cde in underlying model. A numeric vector of length 1. Mediator level at which controlled direct effect is evaluated at.
Covariate_NaturalDirectIndirect	c_cond in underlying model. A numeric vector of the same length as cvar. Covariate levels at which natural direct and indirect effects are evaluated at.
MediatorTargetType	mreg in underlying model. A character vector of length 1. Mediator regression type: "linear" or "logistic".
OutcomeTargetType	yreg in underlying model. A character vector of length 1. Outcome regression type: "linear", "logistic", "loglinear", "poisson", "negbin", "survCox", "survAFT_exp", or "survAFT_weibull".
TreatmentMediatorInteraction	interaction in underlying model. A logical vector of length 1. The presence of treatment-mediator interaction in the outcome model. Default to TRUE.
CaseControlSourceData	casecontrol in underlying model. A logical vector of length 1. Default to FALSE. Whether data comes from a case-control study.
RemoveNA	na_omit in underlying model. A logical vector of length 1. Default to FALSE. Whether to remove NAs in the columns of interest before fitting the models.
ConfoundingVariables	cvar in underlying model. A character vector of length > 0. Covariate names. Use NULL if there is no covariate. However, this is a highly suspicious situation. Even if avar is randomized, mvar is not. Thus, there are usually some confounder(s) to account for the common cause structure (confounding) between mvar and yvar.

Value

list with model output object, summary output, effects output, and an effects plot

Author(s)

Adrian Antico

Examples

## Not run: 
library(regmedint) # to load vv2015
data(vv2015)
Output <- AutoQuant::CausalMediation(
  data = vv2015,
  OutcomeTargetVariable = 'y',            # yvar char length = 0
  TreatmentVariable = "x",                # avar char length = 0 (binary)
  MediatorVariable = "m",                 # mvar char length = 0 (binary)
  Covariates = "c",                       # cvar char length > 0
  MM_TreatmentCovariates = NULL,          # emm_ac_mreg = NULL char length > 0
  OM_TreatmentCovariates = NULL,          # emm_ac_yreg = NULL char length > 0
  OM_MediatorCovariates = NULL,           # emm_mc_yreg = NULL char length > 0
  SurvivalEventVariable = "event",        # eventvar char length = 0
  UnTreated_ReferenceIndicator = 0,       # ao num length = 1
  Treated_ReferenceIndicator = 1,         # a1 num length = 1
  Mediator_ControlDirectEffectLevel = 1,  # m_cde num length = 1
  Covariate_NaturalDirectIndirect = 3,    # c_cond; same length as Covariates num length = length(Covariates)
  MediatorTargetType = 'logistic',        # mreg "linear" or "logistic",
  OutcomeTargetType = 'survAFT_weibull',  # yreg "linear", "logistic", "loglinear", "poisson", "negbin", "survCox", "survAFT_exp", or "survAFT_weibull"
  TreatmentMediatorInteraction = TRUE,    # interaction = TRUE,
  CaseControlSourceData = FALSE,          # casecontrol = FALSE,
  RemoveNA = FALSE)

# data = vv2015
# OutcomeTargetVariable = 'y'
# TreatmentVariable = "x"
# MediatorVariable = "m"
# Covariates = "c"
# MM_TreatmentCovariates = NULL
# OM_TreatmentCovariates = NULL
# OM_MediatorCovariates = NULL
# SurvivalEventVariable = "event"
# UnTreated_ReferenceIndicator = 0
# Treated_ReferenceIndicator = 1
# Mediator_ControlDirectEffectLevel = 1
# Covariate_NaturalDirectIndirect = 3
# MediatorTargetType = 'logistic'
# OutcomeTargetType = 'survAFT_weibull'
# TreatmentMediatorInteraction = TRUE
# CaseControlSourceData = FALSE
# RemoveNA = FALSE

## End(Not run)

AdrianAntico/RemixAutoML documentation built on Feb. 3, 2024, 3:32 a.m.