rumr: Regression Uncertainty Modeler

In dsnavega/rumr: Regression Uncertainty Modeler with R

Regression Uncertainty Modeler

Description

Regression Uncertainty Modeler

Usage

rumr(
  known,
  predicted,
  type,
  alpha = 0.05,
  interval = NULL,
  delta = 0.25,
  signed = F,
  exponent = 1
)

Arguments

known	a numeric vector. Known values used to train or fit a statitical or machine learning regression model
predicted	a numeric vector. Predicted values obtain from a regression model. Ideally this vector should be obtained from cross-validation.
type	a character. The type of uncertainty model to fit. See Details
alpha	uncertainty level. Used when type = "local".
interval	a numeric vector. This vector should have two elements with the minimum and maximum values with the bounds for the regression function. Default is NULL, which means that the values are inferred from known vector. See Details
delta	a numeric value. Controls the extrapolation outside of known values. Must be between 0 and 1. Default is 0.25. See Details.
signed	a logical. Defines if conformal prediction should be signed (assymetrical) or unsigned (symmetrical) prediction intervals around the predicted value. See Details.
exponent	a numeric. Non-linear variance estimation

Details

TODO

Value

a rumr class object.

Author(s)

David Senhora Navega

References

TODO

dsnavega/rumr documentation built on Sept. 15, 2022, 2:10 a.m.