errint (version 1.0)

error_interval: Error Intervals

Description

error_interval creates an object of class c("error_interval", "list").

as.error_interval attempts to coerce its argument x into an object of class c("error_interval", class(x)). If this is not possible x is returned unchanged.

is.error_interval returns TRUE if x is an R object with "error_interval" as one of its classes. It returns FALSE otherwise.

Usage

error_interval(phi, s = 0.05, dist = "n", tol = 10^-6, ...)
as.error_interval(x)
is.error_interval(x)

Arguments

phi

a vector with residual values used to compute the error interval.

s

confidence level, e,g. s=0.05 for the standard 95 percent confidence interval.

dist

assumed distribution for the noise in the data.

tol

used to normalize residual values to (0,1) when beta is the assumed distribution. The formula used is abs(phi)/(max(abs(phi))+tol).

...

additional arguments to be passed to the low level error_interval building functions (see below).

x

an R object.

Value

error_interval returns an object of class c("error_interval","list") with information regarding the error intervals built.as.error_interval returns an object of class c("error_interval",class(x)) with information contained in x if possible. Returns x otherwise.is.error_intervalreturns TRUE if x is an R object with "error_interval" as one of its classes. FALSE otherwise.

References

Link to the scientific paper

Prada, Jesus, and Jose Ramon Dorronsoro. "SVRs and Uncertainty Estimates in Wind Energy Prediction." Advances in Computational Intelligence. Springer International Publishing, 2015. 564-577,

with theoretical background for this package is provided below.

http://link.springer.com/chapter/10.1007/978-3-319-19222-2_47

Examples

error_interval(rnorm(10))

error_interval(rnorm(10),s=0.1,dist="lm")


l<-list(min=-1,max=1,err=0.05,s=0.1,dist="n",phi=rnorm(10))
as.error_interval(l)

v<-c("a","b")
as.error_interval(v)


l<-list(min=-1,max=1,err=0.05,s=0.1,dist="n",phi=rnorm(10))
is.error_interval(l)
res<-as.error_interval(l)
is.error_interval(res)