Get the best model from a pipeline
Value
The best tidylearn_model object from the pipeline,
selected by the metric specified in evaluation$best_metric.
The model was fitted on the preprocessed training data, so under the
default preprocessing it expects predictors imputed and standardised
the same way. Predict through tl_predict_pipeline, which
replays that preprocessing on raw rows; predict() on the model
itself reads raw values as if they were already standardised, and
returns wrong predictions without a warning.
Examples
# \donttest{
pipe <- tl_pipeline(iris, Species ~ .,
models = list(tree = list(method = "tree")),
evaluation = list(metrics = "accuracy", validation = "cv",
cv_folds = 2, best_metric = "accuracy"))
pipe <- tl_run_pipeline(pipe, verbose = FALSE)
best <- tl_get_best_model(pipe)
best$spec$method
#> [1] "tree"
# Predict through the pipeline, which preprocesses the new rows first
tl_predict_pipeline(pipe, iris[c(1, 51, 101), ])
#> # A tibble: 3 × 1
#> .pred
#> <fct>
#> 1 setosa
#> 2 versicolor
#> 3 virginica
# }