predict.FlexCoDEKernel: Evaluates the estimated density of new observations (testing...

In rizbicki/FlexCoDE: Fits Flexible Conditional Estimator via Regression

Description

Evaluates the estimated density of new observations (testing points) of a "FlexCoDE" object

Usage

## S3 method for class 'FlexCoDEKernel'
predict(objectCDE, kernelNewTrain, B = 1000, predictionBandProb = FALSE)

Arguments

objectCDE	Object of the class "FlexCoDE", typically fitted used fitFlexCoDE beforehand
B	Number of point where f(z|x) will be evaluated (on the z scale). This will be equally spaced between zMin and zMax
xNew	Matrix with nTest rows and same number of columns as xTrain, containing x's for which the estimates are desired.

Value

The return value is an object with the following components

z	Points where the density was evaluate
CDE	Matrix with value of the density at points z. Each row corresponds to a different observation x (i-th row of CDE corresponds to i-th row of xTest).
th	(If predictionBandProb is not FALSE) Threshold values for each estimated density. The region where estimated densities are above these values have the approximate coverage probability desired. See plot.FlexCoDE for ploting these regions.

Examples

# See \code{\link{fitFlexCoDE}}

rizbicki/FlexCoDE documentation built on Feb. 10, 2022, 3:14 p.m.