Package ai.djl.basicdataset.tabular

Class TabularDataset.BaseBuilder<T extends TabularDataset.BaseBuilder<T>>

java.lang.Object

ai.djl.training.dataset.RandomAccessDataset.BaseBuilder<T>

ai.djl.basicdataset.tabular.TabularDataset.BaseBuilder<T>

Type Parameters:

T - the builder type

Direct Known Subclasses:

CsvDataset.CsvBuilder

Enclosing class:

TabularDataset

public abstract static class TabularDataset.BaseBuilder<T extends TabularDataset.BaseBuilder<T>>
extends ai.djl.training.dataset.RandomAccessDataset.BaseBuilder<T>

Used to build a TabularDataset.

Field Summary

Fields

Modifier and Type	Field	Description
protected boolean	allowNoLabels
protected List<Feature>	features
protected List<Feature>	labels

Fields inherited from class ai.djl.training.dataset.RandomAccessDataset.BaseBuilder

dataBatchifier, device, labelBatchifier, limit, pipeline, prefetchNumber, sampler, targetPipeline

Constructor Summary

Constructors

Modifier	Constructor	Description
protected	BaseBuilder()

Method Summary

Modifier and Type	Method	Description
T	addCategoricalFeature(String name)	Adds a categorical feature to the feature set.
T	addCategoricalFeature(String name, boolean onehotEncode)	Adds a categorical feature to the feature set.
T	addCategoricalFeature(String name, Map<String,Integer> map, boolean onehotEncode)	Adds a categorical feature to the feature set with specified mapping.
T	addCategoricalLabel(String name)	Adds a categorical feature to the label set.
T	addCategoricalLabel(String name, boolean onehotEncode)	Adds a categorical feature to the label set.
T	addCategoricalLabel(String name, Map<String,Integer> map, boolean onehotEncode)	Adds a categorical feature to the feature set with specified mapping.
T	addFeature(Feature... features)	Adds the features to the feature set.
T	addLabel(Feature... labels)	Adds the features to the label set.
T	addNumericFeature(String name)	Adds a numeric feature to the feature set.
T	addNumericFeature(String name, boolean normalize)	Adds a numeric feature to the feature set.
T	addNumericLabel(String name)	Adds a number feature to the label set.
T	addNumericLabel(String name, boolean normalize)	Adds a number feature to the label set.
T	noLabels()	Indicates the dataset should not have any labels.

Methods inherited from class ai.djl.training.dataset.RandomAccessDataset.BaseBuilder

addTargetTransform, addTransform, getSampler, optDataBatchifier, optDevice, optLabelBatchifier, optLimit, optPipeline, optPrefetchNumber, optTargetPipeline, self, setSampling, setSampling, setSampling

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Details

features

protected List<Feature> features

labels

protected List<Feature> labels

allowNoLabels

protected boolean allowNoLabels

Constructor Details

BaseBuilder

protected BaseBuilder()

Method Details

addFeature

public T addFeature(Feature... features)

Adds the features to the feature set.

Parameters:

features - the features

Returns:

this builder

addNumericFeature

public T addNumericFeature(String name)

Adds a numeric feature to the feature set.

Parameters:

name - the feature name

Returns:

this builder

addNumericFeature

public T addNumericFeature(String name,
 boolean normalize)

Adds a numeric feature to the feature set.

Parameters:

name - the feature name

normalize - true to normalize the column

Returns:

this builder

addCategoricalFeature

public T addCategoricalFeature(String name)

Adds a categorical feature to the feature set.

Parameters:

name - the feature name

Returns:

this builder

addCategoricalFeature

public T addCategoricalFeature(String name,
 boolean onehotEncode)

Adds a categorical feature to the feature set.

Parameters:

name - the feature name

onehotEncode - true to use onehot encode

Returns:

this builder

addCategoricalFeature

public T addCategoricalFeature(String name,
 Map<String,Integer> map,
 boolean onehotEncode)

Adds a categorical feature to the feature set with specified mapping.

Parameters:

name - the feature name

map - a map contains categorical value maps to index

onehotEncode - true to use onehot encode

Returns:

this builder

addLabel

public T addLabel(Feature... labels)

Adds the features to the label set.

Parameters:

labels - the labels

Returns:

this builder

addNumericLabel

public T addNumericLabel(String name)

Adds a number feature to the label set.

Parameters:

name - the label name

Returns:

this builder

addNumericLabel

public T addNumericLabel(String name,
 boolean normalize)

Adds a number feature to the label set.

Parameters:

name - the label name

normalize - true to normalize the column

Returns:

this builder

addCategoricalLabel

public T addCategoricalLabel(String name)

Adds a categorical feature to the label set.

Parameters:

name - the feature name

Returns:

this builder

addCategoricalLabel

public T addCategoricalLabel(String name,
 boolean onehotEncode)

Adds a categorical feature to the label set.

Parameters:

name - the feature name

onehotEncode - true if use onehot encode

Returns:

this builder

addCategoricalLabel

public T addCategoricalLabel(String name,
 Map<String,Integer> map,
 boolean onehotEncode)

Adds a categorical feature to the feature set with specified mapping.

Parameters:

name - the feature name

map - a map contains categorical value maps to index

onehotEncode - true if use onehot encode

Returns:

this builder

noLabels

public T noLabels()

Indicates the dataset should not have any labels.

Returns:

this builder