Deviance

Calculates the deviance.

Usage

.deviance(y, y_hat, family)

Arguments

y

response: numeric vector of length \(n\)

y_hat

predicted response: numeric vector of length \(n\)

family

character

Value

Returns the deviance (a numeric scalar).

Examples

n <- 10

y <- rnorm(n)
y_hat <- rnorm(n)
corila:::.deviance(y = y , y_hat = y_hat, family = "gaussian")
#> [1] 1.41056

y <- rbinom(n = n, size = 1, prob = 0.5)
y_hat <- runif(n)
corila:::.deviance(y = y , y_hat = y_hat, family = "binomial")
#> [1] 1.093534

y <- rpois(n = n, lambda = 4)
y_hat <- rexp(n)
corila:::.deviance(y = y , y_hat = y_hat, family = "poisson")
#> [1] 6.062777