| Type: | Package |
| Title: | A Comprehensive Toolkit for Environmental Mixtures Analysis ('CompMix') |
| Version: | 0.1.0 |
| Description: | Quantitative characterization of the health impacts associated with exposure to chemical mixtures has received considerable attention in current environmental and epidemiological studies. 'CompMix' package allows practitioners to estimate the health impacts from exposure to chemical mixtures data through various statistical approaches, including Lasso, Elastic net, Bayeisan kernel machine regression (BKMR), hierNet, Quantile g-computation, Weighted quantile sum (WQS) and Random forest. Hao W, Cathey A, Aung M, Boss J, Meeker J, Mukherjee B. (2024) "Statistical methods for chemical mixtures: a practitioners guide". <doi:10.1101/2024.03.03.24303677>. |
| License: | GPL-3 |
| Imports: | Matrix, mvtnorm, gglasso, higlasso, hierNet, glmnet, SuperLearner, bkmr, qgcomp, gWQS, pROC, randomForest, devtools |
| Encoding: | UTF-8 |
| RoxygenNote: | 7.2.3 |
| NeedsCompilation: | no |
| Packaged: | 2024-05-21 03:18:28 UTC; weiha |
| Author: | Wei Hao [aut, cre] |
| Maintainer: | Wei Hao <weihao@umich.edu> |
| Repository: | CRAN |
| Date/Publication: | 2024-05-22 12:10:02 UTC |
A comprehensive toolkit for environmental mixtures analysis
Description
A comprehensive toolkit for environmental mixtures analysis
Usage
Comp.Mix(
y,
x,
z = NULL,
y.type,
test.pct = 0.5,
var.select = NULL,
interaction = NULL,
interaction.exp.cov = NULL,
covariates.forcein = NULL,
bkmr.pip = 0.5,
bkmr.iter = 500,
formula = NULL,
expnms = NULL,
seed = 1234,
verbose = TRUE
)
Arguments
y |
A vector of either continuous or binary values to indicate the health outcome |
x |
A matrix of numeric values to indicate the chemical mixtures |
z |
A matrix of numeric values to indicate the covariates |
y.type |
A character value of either "continuous" or "binary" |
test.pct |
A numeric scalar between 0 and 1 to indicate the proportion allocated as test samples |
var.select |
A logical value to indicate whether to perform variable selection |
interaction |
A logical value (TRUE/FALSE) to indicate whether to include pairwise interaction terms between all the chemical mixtures x |
interaction.exp.cov |
A logical value (TRUE/FALSE) to indicate whether to include pairwise interaction terms between all the chemical mixtures x and covariates z. If interaction.exp.cov=TRUE, interaction=TURE or interaction=FALSE will be ignored |
covariates.forcein |
A logical value (TRUE/FALSE) to indicate whether to force in any covariates |
bkmr.pip |
A numeric scalar between 0 and 1 to indicate the cutoff for the posterior inclusion probability in BKMR |
bkmr.iter |
A positive integer to indicate the number of MCMC iterations for bkmr |
formula |
the formula for qgcomp and wqs |
expnms |
a vector of characters for names of exposure variables |
seed |
an integer value for seed |
verbose |
a logical value to show information |
Value
A list object which may contain up to 8 cases
- Case 1
variable selection on main effects for exposures and confounders
Each case may contain some of the following elements
- betaest
a numeric vector of coeffcients for the exposures
- z_betaest
a numeric vector of coeffcients for the covariates
- sse
A positive scalar to indicate sum of squares error
- corr
A numeric scalar between -1 and 1 to indicate correlation coefficient
Author(s)
Wei Hao <weihao@umich.edu>
Examples
dat <- lmi_simul_dat(n=1000,p=20,q=5,
block_idx=c(1,1,2,2,3,1,1,1,1,1,2,2,2,2,3,3,3,3,3,3),
within_rho=0.6,btw_rho=0.1,R2=0.2,
effect_size=1,effect_size_i=1,
cancel_effect = FALSE)
#Example 1: The users would like to perform variable selections
#on main effects of exposures and covariates, and outcome, exposures and
#covariates are entered. For any individual interactions that the users would
#like to include in the models, they can add those into the covariate z.
res_ex1 <- Comp.Mix(y.type="continuous",y=dat$y, x=dat$x, z=dat$z, test.pct=0.5,
var.select = TRUE, interaction = FALSE, interaction.exp.cov = FALSE,
covariates.forcein = FALSE,
bkmr.pip=0.5, seed=2023)
Simulate data from linear model with interactions
Description
Simulate data from linear model with interactions
Usage
lmi_simul_dat(
n,
p,
q,
block_idx = c(1, 1, 2, 2, 3, 1, 1, 1, 1, 1, 2, 2, 2, 2, 3, 3, 3, 3, 3, 3),
sigma2_x = 1,
within_rho = 0.6,
btw_rho = 0.2,
R2 = 0.8,
effect_size = 1,
effect_size_i = 1,
cancel_effect = TRUE
)
Arguments
n |
a positive integer to indicate sample size |
p |
a positive integer to specify the number of exposures |
q |
a positive integer to specify the number of non-zero effects |
block_idx |
a vector of positive integers to indicate the block IDs. The length of the vector is p. |
sigma2_x |
a positive numeric scalar for variance of the covariates |
within_rho |
a numeric scalar between 0 and 1 for the within block correlation |
btw_rho |
a numeric scalar between 0 and 1 for the between block correlation |
R2 |
a numeric scalar for R-squared |
effect_size |
a numeric scalar for effect size for main effect |
effect_size_i |
a numeric scalar for effect size for interaction effect |
cancel_effect |
a logic value to indicate whether there is effect cancelation |
Value
a list object of the following
- x
covariate matrix of dimension n by p
- n
sample size
- p
number of covariates
- sigma2_x
variance
- within_rho
within block correlation
- btw_rho
between block correaltion
- block_idx
block indices
Author(s)
Wei Hao <weihao@umich.edu>
Simulate covariate matrix with block structure
Description
Simulate covariate matrix with block structure
Usage
simul_x_block(n, p, block_idx, sigma2_x = 1, within_rho = 0.6, btw_rho = 0.2)
Arguments
n |
a positive integer to indicate sample size |
p |
a positive integer to specify the number of covariates |
block_idx |
a vector of positive integers to indicate the block IDs. The length of the vector is p. |
sigma2_x |
a positive numeric scalar for variance of the covariates |
within_rho |
a numeric scalar between 0 and 1 for the within block correlation |
btw_rho |
a numeric scalar between 0 and 1 for the between block correlation |
Value
a list object of the following
- x
covariate matrix of dimension n by p
- n
sample size
- p
number of covariates
- sigma2_x
variance
- within_rho
within block correlation
- btw_rho
between block correaltion
- block_idx
block indices
Author(s)
Wei Hao <weihao@umich.edu>
Examples
dat <- simul_x_block(n = 1000, p = 10, block_idx = rep(1:4,length=10))