truncated_normal_cdf {LTFHPlus} | R Documentation |
CDF for truncated normal distribution.
Description
truncated_normal_cdf
computes the cumulative density
function for a truncated normal distribution.
Usage
truncated_normal_cdf(
liability,
lower = stats::qnorm(0.05, lower.tail = FALSE),
upper = Inf
)
Arguments
liability |
A number representing the individual's true underlying liability. |
lower |
A number representing the lower cutoff point for the truncated normal distribution. Defaults to 1.645 (stats::qnorm(0.05, lower.tail = FALSE)). |
upper |
A number representing the upper cutoff point of the truncated normal distribution. Must be greater or equal to lower. Defaults to Inf. |
Details
This function can be used to compute the value of the cumulative density function for a truncated normal distribution given an individual's true underlying liability.
Value
If liability is a number and the lower and upper cutoff points
are numbers satisfying lower <= upper, then truncated_normal_cdf
returns the probability that the liability will take on a value less than
or equal to liability
.
Examples
curve(sapply(liability, truncated_normal_cdf), from = qnorm(0.05, lower.tail = FALSE), to = 3.5,
xname = "liability")