MAZE {MAZE} | R Documentation |
Mediation Analysis for ZEro-inflated mediators
Description
A novel mediation modeling approach to address zero-inflated mediators containing true zeros and false zeros.
Usage
MAZE(
data,
distM = c("zilonm", "zinbm", "zipm"),
K = 1,
selection = "AIC",
X,
M,
Y,
Z = NULL,
XMint = c(TRUE, FALSE),
x1,
x2,
zval = NULL,
mval = 0,
B = 20,
seed = 1,
ncore = 1
)
Arguments
data |
a data frame containing variables: an independent variable |
distM |
a vector with choices of the distribution of mediator to try with. One or more of ' |
K |
a vector with choices of the number of component |
selection |
model selection criterion when more than one model (combination of different values in |
X |
name of the independent variable. Can be continuous or discrete |
M |
name of the mediator variable. Non-negative values |
Y |
name of the outcome variable. Continuous values |
Z |
name(s) of confounder variables (if any) |
XMint |
a logical vector of length 2 indicating whether to include the two exposure-mediator interaction terms between (i) |
x1 |
the first value of independent variable of interest |
x2 |
the second value of independent variable of interest |
zval |
a vector of value(s) of confounders to be conditional on when estimating effects |
mval |
the fixed value of mediator to be conditional on when estimating CDE |
B |
the upper bound value |
seed |
an optional seed number to control randomness |
ncore |
number of cores available for parallel computing |
Details
For an independent variable X
, a zero-inflated mediator M
and a continuous outcome variable Y
, the following regression equation is used to model the association between Y
and (X,M)
:
Y_{xm1_{(m>0)}}=\beta_0+\beta_1m+\beta_2 1_{(m>0)}+\beta_3x+\beta_4x1_{(m>0)}+\beta_5xm+\epsilon
Users can choose to include either one, both, or none of the two exposure-mediator interaction terms between (i) X
and 1_{(M>0)}
and (ii) X
and M
using the argument XMint
.
For mediators, zero-inflated log-normal, zero-inflated negative binomial, and zero-inflated Poisson distributions are considered and can be specified through the argument distM
.
The indirect and direct effects (NIE1, NIE2, NIE, NDE, and CDE) are estimated for X
changing from x1
to x2
. When confounders are present, the conditional effects are estimated given the fixed value zval
.
Value
a list containing:
-
results_effects
: a data frame for the results of estimated effects (NIE1, NIE2, NIE, NDE, and CDE) -
results_parameters
: a data frame for the results of model parameters -
selected_model_name
: a string for the distribution ofM
and number of componentsK
selected in the final mediation model -
BIC
: a numeric value for the BIC of the final mediation model -
AIC
: a numeric value for the AIC of the final mediation model -
models
: a list with all fitted models -
analysis2_out
: a list with output fromanalysis2()
function (used for internal check)
Author(s)
Meilin Jiang meilin.jiang@ufl.edu and Zhigang Li zhigang.li@ufl.edu
Examples
data(zinb10)
maze_out <- MAZE(data = zinb10,
distM = c('zilonm', 'zinbm', 'zipm'), K = 1,
selection = 'AIC',
X = 'X', M = 'Mobs', Y = 'Y', Z = NULL,
XMint = c(TRUE, FALSE),
x1 = 0, x2 = 1, zval = NULL, mval = 0,
B = 20, seed = 1)
## results of selected mediation model
maze_out$results_effects # indirect and direct effects
maze_out$selected_model_name # selected distribution of the mediator and number of components K
maze_out$results_parameters # model parameters
maze_out$BIC; maze_out$AIC # BIC and AIC of the selected mediation model