pif.confidence.loglinear: Confidence intervals for the Potencial Impact Fraction, using...

In pifpaf: Potential Impact Fraction and Population Attributable Fraction for Cross-Sectional Data

Description

Confidence intervals for the Potencial Impact Fraction for relative risk inyective functions, the pif is inyective, and intervals can be calculated for log(pif), and then transformed to pif CI.

Usage

pif.confidence.loglinear(X, thetahat, thetavar, rr, cft = NA,
  weights = rep(1/nrow(as.matrix(X)), nrow(as.matrix(X))), nsim = 100,
  confidence = 95, check_thetas = TRUE, check_exposure = TRUE,
  check_cft = TRUE, is_paf = FALSE)

Arguments

X	Random sample (can be vector or matrix) which includes exposure and covariates.
thetahat	Estimative of theta for the Relative Risk function
thetavar	Estimator of variance of thetahat
rr	Function for Relative Risk which uses parameter theta. The order of the parameters shound be rr(X, theta). **Optional**
cft	Function cft(X) for counterfactual. Leave empty for the Population Attributable Fraction PAF where counterfactual is 0 exposure.
weights	Survey weights for the random sample X.
nsim	Number of simulations for estimation of variance.
confidence	Confidence level % (default 95)
check_thetas	Check that theta parameters are correctly inputed
check_exposure	Check that exposure X is positive and numeric
check_cft	Check if counterfactual function cft reduces exposure.
is_paf	Force evaluation of paf

Author(s)

Rodrigo Zepeda-Tello [e-mail]

Dalia Camacho-Garc<c3><ad>a-Forment<c3><ad> [e-mail]

Examples

#Example 1: Exponential Relative risk
#--------------------------------------------
set.seed(18427)
X        <- as.data.frame(rnorm(100,5,1))
thetahat <- 0.4
thetavar <- 0.1
cft      <- function(X){sqrt(X)}
pif.confidence.loglinear(X, thetahat, thetavar, rr = function(X, theta){exp(theta*X)}, cft)

#Same example with linear counterfactual
a    <- 0.5
cft  <- function(X){a*X}
pif.confidence.loglinear(X, thetahat, thetavar, function(X, theta){exp(theta*X)})

#' #Example 2: Multivariate Relative Risk
#--------------------------------------------
set.seed(18427)
X1        <- rnorm(100, 4,0.01)
X2        <- runif(100,0.4,2)
X         <- as.data.frame(cbind(X1,X2))
thetahat  <- c(0.12, 0.03)
thetavar  <- matrix(c(0.01, 0, 0, 0.04), byrow = TRUE, nrow = 2)
rr        <- function(X, theta){
           .X <- as.matrix(X, ncol = 2)
           exp(theta[1]*.X[,1] + theta[2]*.X[,2])
           }
pif.confidence.loglinear(X, thetahat, thetavar, rr) 

pifpaf documentation built on May 1, 2019, 9:11 p.m.