Regmedint
R implementation of effect measure modification-extended regression-based closed-formula causal mediation analysis
Install / Use
/learn @kaz-yos/RegmedintREADME
regmedint <img src="man/figures/hex.png" align="right" height="140"/>
This is an extension of the regression-based causal mediation analysis first proposed by Valeri and VanderWeele (2013) and Valeri and VanderWeele (2015). The current version supports including effect measure modification by covariates (treatment-covariate and mediator-covariate product terms in mediator and outcome regression models). It also accommodates the original ‘SAS’ macro (can be found at Dr. VanderWeele’s Tools and Tutorials) and PROC CAUSALMED procedure in ‘SAS’ when there is no effect measure modification. Linear and logistic models are supported for the mediator model. Linear, logistic, loglinear, Poisson, negative binomial, Cox, and accelerated failure time (exponential and Weibull) models are supported for the outcome model.
To cite this software, please use: regmedint (v1.0.0; Yoshida, Li, & Mathur, 2021)
<!-- cite the r package: https://easystats.github.io/report/articles/cite_packages.html -->Implemented models
The following grid of models are implemented. yreg refers to the
outcome model and mreg refers to the mediator model.
<sup>1</sup> Approximation depends on the rare event assumptions.
<sup>2</sup> Implemented as a modified Poisson model (log link with robust variance) as in Z2004.
See the corresponding vignettes (Articles on the package website) for how to perform bootstrapping and multiple imputation.
Installation
For the developmental version on Github, use the following commands to install the package.
# install.packages("devtools") # If you do not have devtools already.
devtools::install_github("kaz-yos/regmedint")
##
## checking for file ‘/private/var/folders/5m/w191nn3d52bc91mq5_jjgw880000gn/T/RtmpTDYzP5/remotesf00c1976f8fd/kaz-yos-regmedint-43d42e2/DESCRIPTION’ ... ✔ checking for file ‘/private/var/folders/5m/w191nn3d52bc91mq5_jjgw880000gn/T/RtmpTDYzP5/remotesf00c1976f8fd/kaz-yos-regmedint-43d42e2/DESCRIPTION’
## ─ preparing ‘regmedint’:
## checking DESCRIPTION meta-information ... ✔ checking DESCRIPTION meta-information
## ─ checking for LF line-endings in source and make files and shell scripts
## ─ checking for empty or unneeded directories
## Removed empty directory ‘regmedint/man/figures’
## ─ building ‘regmedint_1.0.0.tar.gz’
##
##
The CRAN version can be installed as follows.
install.packages("regmedint")
Data Example
We use VV2015 dataset for demonstration.
library(regmedint)
data(vv2015)
regmedint() to fit models
The regmedint function is the user interface for constructing a result
object of class regmedint. The interface is similar to the original
SAS macro. For survival outcomes, the indicator variable is an event
indicator (1 for event, 0 for censoring). c_cond vector is required be
specified. This vector is the vector of covariate values at which the
conditional effects are evaluated at.
- When there is no effect measure modification by covariates,
emm_ac_mreg = NULL,emm_ac_yreg = NULL,emm_mc_yreg = NULL.
regmedint_obj1 <- regmedint(data = vv2015,
## Variables
yvar = "y",
avar = "x",
mvar = "m",
cvar = c("c"),
eventvar = "event",
## Values at which effects are evaluated
a0 = 0,
a1 = 1,
m_cde = 1,
c_cond = 3,
## Model types
mreg = "logistic",
yreg = "survAFT_weibull",
## Additional specification
interaction = TRUE,
casecontrol = FALSE)
summary(regmedint_obj1)
## ### Mediator model
##
## Call:
## glm(formula = m ~ x + c, family = binomial(link = "logit"), data = data)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.5143 -1.1765 0.9177 1.1133 1.4602
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -0.3545 0.3252 -1.090 0.276
## x 0.3842 0.4165 0.922 0.356
## c 0.2694 0.2058 1.309 0.191
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 138.59 on 99 degrees of freedom
## Residual deviance: 136.08 on 97 degrees of freedom
## AIC: 142.08
##
## Number of Fisher Scoring iterations: 4
##
## ### Outcome model
##
## Call:
## survival::survreg(formula = Surv(y, event) ~ x + m + x:m + c,
## data = data, dist = "weibull")
## Value Std. Error z p
## (Intercept) -1.0424 0.1903 -5.48 0.000000043
## x 0.4408 0.3008 1.47 0.14
## m 0.0905 0.2683 0.34 0.74
## c -0.0669 0.0915 -0.73 0.46
## x:m 0.1003 0.4207 0.24 0.81
## Log(scale) -0.0347 0.0810 -0.43 0.67
##
## Scale= 0.966
##
## Weibull distribution
## Loglik(model)= -11.4 Loglik(intercept only)= -14.5
## Chisq= 6.31 on 4 degrees of freedom, p= 0.18
## Number of Newton-Raphson Iterations: 5
## n= 100
##
## ### Mediation analysis
## est se Z p lower upper
## cde 0.541070807 0.29422958 1.8389409 0.06592388 -0.03560858 1.11775019
## pnde 0.505391952 0.21797147 2.3186151 0.02041591 0.07817572 0.93260819
## tnie 0.015988820 0.03171597 0.5041252 0.61417338 -0.04617334 0.07815098
## tnde 0.513662425 0.22946248 2.2385465 0.02518544 0.06392423 0.96340062
## pnie 0.007718348 0.02398457 0.3218047 0.74760066 -0.03929055 0.05472725
## te 0.521380773 0.22427066 2.3247837 0.02008353 0.08181835 0.96094319
## pm 0.039039346 0.07444080 0.5244348 0.59997616 -0.10686194 0.18494063
##
## Evaluated at:
## avar: x
## a1 (intervened value of avar) = 1
## a0 (reference value of avar) = 0
## mvar: m
## m_cde (intervend value of mvar for cde) = 1
## cvar: c
## c_cond (covariate vector value) = 3
##
## Note that effect estimates can vary over m_cde and c_cond values when interaction = TRUE.
- When there is effect measure modification by covariates,
emm_ac_mreg,emm_ac_yregandemm_mc_yregcan take a sub-vector of covariates incvar.
regmedint_obj2 <- regmedint(data = vv2015,
## Variables
yvar = "y",
avar = "x",
mvar = "m",
cvar = c("c"),
emm_ac_mreg = c("c"),
emm_ac_yreg = c("c"),
emm_mc_yreg = c("c"),
eventvar = "event",
## Values at which effects are evaluated
a0 = 0,
a1 = 1,
m_cde = 1,
c_cond = 3,
## Model types
mreg = "logistic",
Related Skills
node-connect
350.1kDiagnose OpenClaw node connection and pairing failures for Android, iOS, and macOS companion apps
frontend-design
109.9kCreate distinctive, production-grade frontend interfaces with high design quality. Use this skill when the user asks to build web components, pages, or applications. Generates creative, polished code that avoids generic AI aesthetics.
openai-whisper-api
350.1kTranscribe audio via OpenAI Audio Transcriptions API (Whisper).
qqbot-media
350.1kQQBot 富媒体收发能力。使用 <qqmedia> 标签，系统根据文件扩展名自动识别类型（图片/语音/视频/文件）。
