Type: Package
Title: Compositional Mediation Model
Version: 1.0.0
Description: A compositional mediation model for continuous outcome and binary outcomes to deal with mediators that are compositional data. Lin, Ziqiang et al. (2022) <doi:10.1016/j.jad.2021.12.019>.
Depends: R (≥ 3.5.0)
License: GPL-3
Encoding: UTF-8
Imports: fastDummies,survey,robCompositions,ggplot2,forcats,dplyr
LazyData: true
RoxygenNote: 7.1.2
NeedsCompilation: no
Packaged: 2022-10-14 17:05:49 UTC; linzi
Author: Ziqiang Lin [aut, cre], Jinqun Cheng [aut], Qiaoxuan Lin [aut], Wayne Lawrence [aut], Wangjian Zhang [aut], Yanhui Gao [aut]
Maintainer: Ziqiang Lin <linziqiang0314@gmail.com>
Repository: CRAN
Date/Publication: 2022-10-17 06:35:19 UTC

Test Data

Description

Contains artificial 100 samples with a continuous outcome variable y, a continuous treatment tr, 20 compositional mediators M and 2 covariates X. The true direct and indirect effects of treatment on the outcome both are 1.00. The true component-wise indirect effects (M1-M20) are 0.693, -0.425, 0.135, -0.057, -0.268, 0.970, -0.843, 0.805, 0.000, 0.000, 0.000, 0.000, 0.000, 0.000, 0.000, 0.000, 0.000, 0.000, 0.000, 0.000.

Usage

data(CMM_test_data)

Compositional Mediation Model

Description

A compositional mediation model for continuous outcome and binary outcomes to deal with mediators that are compositional data

Usage

CMMs(data,outcome,med,pred,cov_con=NULL,cov_cat=NULL,
           weight=NULL,family="identity",boot=5000)

Arguments

data

an input dataframe

outcome

column number that locate continuous or binary outcome variable in data

med

a vector of column numbers that locate the compositional mediators in data

pred

column number that locate continuous or binary exposure in data

cov_con

a vector of column numbers that locate the continuous covariates in data (default NULL).

cov_cat

a vector of column numbers that locate the categorical covariates in data (default NULL).

weight

column number that locate weights in data (default NULL).

family

If your outcome variable is continuous, then family="identity"; if your outcome variable is binary, then family="logistic" (default "identity")

boot

Number of bootstrap (default 5000)

Details

This code can be used to model with a situation when the mediators are compositional data.

Value

An object of class CMM, which is a list with the following components:

Indirect.effect

Indirect effects of exposure on an outcome variable (with 95% bootstrap confidence intervals)

Direct.effect

Direct effects of exposure on an outcome variable (with 95% bootstrap confidence intervals)

Total.effect

Total effects of exposure on an outcome variable (with 95% bootstrap confidence intervals)

Mediation.effect.plot

A plot shows mediation effect of exposure on an outcome variables (mediation effect with with 95% bootstrap confidence intervals)

Relative.Effects.plot

A plot shows relative effect of exposure on an outcome variables (relative effect with with 95% bootstrap confidence intervals)

References

Lin Z, Zhu S, Cheng J, Lin Q, Lawrence WR, Zhang W, Huang Y, Chen Y, Gao Y. The mediating effect of engagement in physical activity over a 24-hour period on chronic disease and depression: using compositional mediation model. J Affect Disord. 2021 Dec 10:S0165-0327(21)01337-9. doi: 10.1016/j.jad.2021.12.019.

Examples

data(CMM_test_data)
result=CMMs(CMM_test_data,1,3:22,2,cov_con=23:24,cov_cat=NULL,weight=NULL,boot=100)