feature_spec: Creates a feature specification.

Description Usage Arguments Details Value See Also Examples

View source: R/feature_spec.R

Description

Used to create initialize a feature columns specification.

Usage

1
feature_spec(dataset, x, y = NULL)

Arguments

dataset

A TensorFlow dataset.

x

Features to include can use tidyselect::select_helpers() or a formula.

y

(Optional) The response variable. Can also be specified using a formula in the x argument.

Details

After creating the feature_spec object you can add steps using the step functions.

Value

a FeatureSpec object.

See Also

Other Feature Spec Functions: dataset_use_spec(), fit.FeatureSpec(), step_bucketized_column(), step_categorical_column_with_hash_bucket(), step_categorical_column_with_identity(), step_categorical_column_with_vocabulary_file(), step_categorical_column_with_vocabulary_list(), step_crossed_column(), step_embedding_column(), step_indicator_column(), step_numeric_column(), step_remove_column(), step_shared_embeddings_column(), steps

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
## Not run: 
library(tfdatasets)
data(hearts)
hearts <- tensor_slices_dataset(hearts) %>% dataset_batch(32)

# use the formula interface
spec <- feature_spec(hearts, target ~ .)

# select using `tidyselect` helpers
spec <- feature_spec(hearts, x = c(thal, age), y = target)

## End(Not run)

tfdatasets documentation built on Nov. 10, 2021, 1:07 a.m.