Description Usage Arguments Note Author(s) See Also Examples
Function that calculates the potential impact fraction pif
via the empirical method. That is: for a random sample X
, a relative
risk function rr(X, thetahat)
with parameters thetahat
the
empirical estimator is given by:
PIF = 1 - ∑ rr(cft(X_i); θ)/∑ rr(X_i; θ)
1 2 3 |
X |
Random sample ( |
thetahat |
Estimator ( |
rr |
Function for Relative Risk which uses parameter **Optional** |
cft |
Function |
weights |
Normalized survey |
check_exposure |
Check that exposure |
check_rr |
Check that Relative Risk function |
check_integrals |
Check that counterfactual and relative risk's expected values are well defined for this scenario. |
is_paf |
Boolean forcing evaluation of |
The empirical method converges for relative risk rr
functions
that are Lipschitz, convex or concave on thetahat
. For stranger
functions use pif.kernel
.
Rodrigo Zepeda-Tello rzepeda17@gmail.com
Dalia Camacho-García-Formentí daliaf172@gmail.com
pif
which is a wrapper for all pif methods
(pif.empirical
, pif.approximate
,
pif.kernel
).
For estimation of the Population Attributable Fraction see
paf
.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 | #Example 1: Relative risk given by exponential
#--------------------------------------------
set.seed(18427)
X <- data.frame(rnorm(100,3,.5))
thetahat <- 0.12
rr <- function(X, theta){exp(theta*X)}
pif.empirical(X, thetahat, rr, cft = function(X){ 0.5*X })
#Without counterfactual estimates PAF
pif.empirical(X, thetahat, rr)
#Example 2: Linear relative risk
#--------------------------------------------
pif.empirical(X, thetahat, rr = function(X, theta){theta*X + 1},
cft = function(X){ 0.5*X })
#Example 3: Multivariate relative risk
#--------------------------------------------
set.seed(18427)
X1 <- rnorm(100,4,1)
X2 <- rnorm(100,2,0.4)
X <- data.frame(cbind(X1,X2))
thetahat <- c(0.12, 0.03)
rr <- function(X, theta){exp(theta[1]*X[,1] + theta[2]*X[,2])}
#Creating a counterfactual. As rr requires a bivariate input, cft should
#return a two-column matrix
cft <- function(X){
cbind(X[,1]/2, 1.1*X[,2])
}
pif.empirical(X, thetahat, rr, cft)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.