Extract residuals

Extract residuals

Usage

# S3 method for class 'brs'
residuals(
  object,
  type = c("response", "pearson", "deviance", "rqr", "weighted", "sweighted"),
  ...
)

Arguments

object

A fitted "betaregscale" object.

type

Residual type. One of "response" (default), "pearson", "deviance", "rqr" (randomized quantile), "weighted", or "sweighted".

...

Currently ignored.

Value

Numeric vector of residuals.

Details

For Pearson residuals the variance formula depends on the reparameterization stored in object$repar:

repar = 1 (precision)

V = mu(1 - mu) / (1 + phi)

repar = 2 (mean-variance)

V = mu(1 - mu) * phi

The weighted and sweighted residuals use the digamma/trigamma formulation from the precision parameterization (repar = 1), so internal conversion is applied when repar != 1.

See also

brs, fitted.brs, plot.brs

Examples

# \donttest{
dat <- data.frame(
  y = c(
    0, 5, 20, 50, 75, 90, 100, 30, 60, 45,
    10, 40, 55, 70, 85, 25, 35, 65, 80, 15
  ),
  x1 = rep(c(1, 2), 10)
)
prep <- brs_prep(dat, ncuts = 100)
#> brs_prep: n = 20 | exact = 0, left = 1, right = 1, interval = 18
fit <- brs(y ~ x1, data = prep)
head(residuals(fit))
#> [1] -0.50871261 -0.40380411 -0.30872261  0.04619589  0.24127739  0.44619589
head(residuals(fit, type = "pearson"))
#>          1          2          3          4          5          6 
#> -1.6029039 -1.2776185 -0.9727549  0.1461618  0.7602416  1.4117443 
# }