normalVar: Compute variance of truncated normal

In selection: Correcting Biased Estimates Under Selection

Description

Given a particular cutoff (x), what is the variance of the new truncated normal?

Usage

normalVar(cutoff, truncMean = NULL, mean = 0, sd = 1)

Arguments

cutoff	The cutoff value for the truncated normal
truncMean	The mean of the truncated normal. If left NULL, it will be estimated.
mean	The mean of the non-truncated distribution (defaults to zero)
sd	The sd of the non-truncated distribution (defaults to 1)

Details

The variance of a pdf (f(x)) is computed by integrating the function x^2*f(x) from -Inf to Inf. To compute a truncated normal, the function remains the same, but the limits change (from -Inf to cutoff).

Value

the variance of the tuncated normal distribution

Author(s)

Dustin Fife

Examples

### compute variance of a distribution cutoff at zero
normalVar(1)
### compare to simulated data
x = rnorm(10000000, 0, 1)
var(x[x>1])

selection documentation built on May 30, 2017, 5:27 a.m.