pif.variance.approximate.linear: Approximate Variance for the Potential Impact Fraction using...

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

Description

Function that calculates approximate variance to the potential impact fraction.

Usage

pif.variance.approximate.linear(X, thetahat, rr, thetavar, Xvar, cft = NA,
  check_thetas = TRUE, check_cft = TRUE, check_xvar = TRUE,
  check_rr = TRUE, check_integrals = TRUE, check_exposure = TRUE,
  deriv.method.args = list(), deriv.method = c("Richardson", "complex"),
  nsim = 1000, is_paf = FALSE)

Arguments

X	Mean value of exposure levels from a cross-sectional random sample.
thetahat	Estimator (vector or matrix) of theta for the Relative Risk function.
rr	Function for Relative Risk which uses parameter theta. The order of the parameters shound be rr(X, theta). **Optional**
thetavar	Estimator of variance of thetahat
Xvar	Variance of exposure levels.
cft	Function cft(X) for counterfactual. Leave empty for the Population Attributable Fraction paf where counterfactual is 0 exposure.
check_thetas	Checks that theta parameters are correctly inputed
check_cft	Check if counterfactual function cft reduces exposure.
check_xvar	Check if it is covariance matrix.
check_rr	Check that Relative Risk function rr equals 1 when evaluated at 0
check_integrals	Check that counterfactual and relative risk's expected values are well defined for this scenario.
check_exposure	Check that exposure X is positive and numeric
deriv.method.args	method.args for hessian.
deriv.method	method for hessian. Don't change this unless you know what you are doing.
nsim	Number of simulations for estimation of variance
is_paf	Force evaluation as paf

Author(s)

Rodrigo Zepeda-Tello [e-mail]

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

See Also

pif.variance.linear for pif variance and pif.confidence for confidence intervals of pif

Examples

## Not run: 
#Example 1: Exponential Relative risk
#--------------------------------------------
set.seed(46987)
rr       <- function(X,theta){exp(X*theta)}
cft      <- function(X){0.5*X}
X        <- rbeta(100, 2, 3)
Xmean    <- data.frame(mean(X))
Xvar     <- var(X)
theta    <- 1.2
thetavar <- 0.15
pif.variance.approximate.linear(Xmean, theta, rr, thetavar, Xvar, cft) 

#Example 2: Multivariate example
#--------------------------------------------
X1       <- rnorm(1000,3,.5)
X2       <- rnorm(1000,4,1)
X        <- data.frame(cbind(X1,X2))
Xmean    <- matrix(colMeans(X), ncol = 2)
Xvar     <- cov(X)
theta    <- c(0.12, 0.17)
thetavar  <- matrix(c(0.001, 0.00001, 0.00001, 0.004), byrow = TRUE, nrow = 2)
rr       <- function(X, theta){exp(theta[1]*X[,1] + theta[2]*X[,2])}
pif.variance.approximate.linear(Xmean, theta, rr, thetavar, Xvar,
cft = function(X){cbind(0.5*X[,1],0.4*X[,2])}, check_integrals = FALSE)

## End(Not run)

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