compete: Returns the winning model

In autovarCore: Automated Vector Autoregression Models and Networks

Description

This function returns the best model as explained in the documentation for the autovar function.

Usage

compete(best, challenger, compare_outliers)

Arguments

best	A model given as a list with at least the properties model_score, nr_dummy_variables, and bucket.
challenger	Another model, also given as a list with properties model_score, nr_dummy_variables, and bucket.
compare_outliers	A boolean. When FALSE, the model comparison does not take the number of dummy variables into account.

Value

This function returns the best model of the two given models.

Examples

model1 <- list(logtransformed = FALSE, lag = 1, nr_dummy_variables = 1,
               model_score = 100, bucket = 0.05)
model2 <- list(logtransformed = FALSE, lag = 2, nr_dummy_variables = 2,
               model_score = 200, bucket = 0.01)
autovarCore:::compete(model1, model2, TRUE)

Example output

$logtransformed
[1] FALSE

$lag
[1] 1

$nr_dummy_variables
[1] 1

$model_score
[1] 100

$bucket
[1] 0.05

autovarCore documentation built on May 2, 2019, 4:01 a.m.