box.dot: Box density ordinate transform

Description Usage Arguments Details Value

View source: R/Performance-metrics.R

Description

Run Box density ordinate transform over multivariate normal model to assess calibration.

Usage

1
box.dot(o, mu, s)

Arguments

o

Vector of observations

mu

Vector of fitted means

s

Fitted covariance matrix

Details

Returns a centre-outward ordering: outliers tend to have low values, inliers tend to have high values. (Tells us nothing about direction of any bias, though)

Value

Ordinate-transformed value


ClairBee/SX.weather documentation built on May 6, 2019, 11:51 a.m.