calfun: Estimating the Calibration Equation

In CVcalibration: Estimation of the Calibration Equation with Error-in Observations

Description

Estimating the calibration equation “y=a+b*x” with error-in observations assuming that the coefficients of the variation of the measurements are constants.

Usage

calfun(x, y, CVx, CVy, lambda0)

Arguments

x	The observed $x$ values
y	The observed $y$ values
CVx	The underlying coefficient of variation of measurement $x$
CVy	The underlying coefficient of variation of measurement $y$
lambda0	The ratio, $CV_y^2/CV_x^2$

Value

result

The estimated regression coefficients, standard error and confidence intervals based on (1) CVx only; (2) CVy only; (3) both CVx and CVy; and (4) the ratio of CVy^2/CVx^2.

Author(s)

Lu Tian, He Qi

Examples

n=100
sigma0=10

beta0=5
beta1=1.2
CVx=0.15
CVy=0.07

lambda0=CVy^2/CVx^2

x0=runif(n, 20, 200)
y0=beta0+beta1*x0+rnorm(n)*sigma0
x=x0+x0*CVx*rnorm(n)
y=y0+y0*CVy*rnorm(n)

fit=calfun(x, y, CVx, CVy, lambda0)
fit

CVcalibration documentation built on May 2, 2019, 2:36 p.m.