Uses of Interface
org.tribuo.DataSource
Packages that use DataSource
Package
Description
Provides the core interfaces and classes for using Tribuo.
Simple data sources for ingesting or aggregating data.
Evaluation base classes, along with code for train/test splits and cross validation.
-
Uses of DataSource in org.tribuo
Subinterfaces of DataSource in org.tribuoModifier and TypeInterfaceDescriptioninterfaceConfigurableDataSource<T extends Output<T>>It's aDataSourcethat's alsoConfigurable.Constructors in org.tribuo with parameters of type DataSourceModifierConstructorDescriptionprotectedDataset(DataSource<T> dataSource) Creates a dataset.ImmutableDataset(DataSource<T> dataSource, FeatureMap featureIDMap, OutputInfo<T> outputIDInfo, boolean dropInvalidExamples) Creates a dataset from a data source.ImmutableDataset(DataSource<T> dataSource, Model<T> model, boolean dropInvalidExamples) Creates a dataset from a data source.MutableDataset(DataSource<T> dataSource) Creates a dataset from a data source. -
Uses of DataSource in org.tribuo.datasource
Classes in org.tribuo.datasource that implement DataSourceModifier and TypeClassDescriptionclassAggregateConfigurableDataSource<T extends Output<T>>Aggregates multipleConfigurableDataSources, usesAggregateDataSource.IterationOrderto control the iteration order.classAggregateDataSource<T extends Output<T>>Aggregates multipleDataSources, usesAggregateDataSource.IterationOrderto control the iteration order.final classIDXDataSource<T extends Output<T>>A DataSource which can read IDX formatted data (i.e., MNIST).final classLibSVMDataSource<T extends Output<T>>A DataSource which can read LibSVM formatted data.classListDataSource<T extends Output<T>>A data source which wraps up a list ofExamples along with theirDataSourceProvenanceand anOutputFactory.Constructor parameters in org.tribuo.datasource with type arguments of type DataSourceModifierConstructorDescriptionAggregateDataSource(List<DataSource<T>> sources) Creates an aggregate data source which will iterate the provided sources in the order of the list (i.e., usingAggregateDataSource.IterationOrder.SEQUENTIAL.AggregateDataSource(List<DataSource<T>> sources, AggregateDataSource.IterationOrder order) Creates an aggregate data source using the supplied sources and iteration order. -
Uses of DataSource in org.tribuo.evaluation
Methods in org.tribuo.evaluation that return DataSourceModifier and TypeMethodDescriptionTrainTestSplitter.getTest()Gets the testing datasource.TrainTestSplitter.getTrain()Gets the training data source.Methods in org.tribuo.evaluation with parameters of type DataSourceModifier and TypeMethodDescriptionfinal EAbstractEvaluator.evaluate(Model<T> model, DataSource<T> datasource) Produces an evaluation for the supplied model and datasource, by callingModel.predict(org.tribuo.Example<T>)to create the predictions, then aggregating the appropriate statistics.Evaluator.evaluate(Model<T> model, DataSource<T> datasource) Evaluates the dataset using the supplied model, returning an immutableEvaluationof the appropriate type.Constructors in org.tribuo.evaluation with parameters of type DataSourceModifierConstructorDescriptionTrainTestSplitter(DataSource<T> data) Creates a splitter that splits a dataset 70/30 train and test using a default seed.TrainTestSplitter(DataSource<T> data, double trainProportion, long seed) Creates a splitter that will split the given data set into a training and testing set.TrainTestSplitter(DataSource<T> data, long seed) Creates a splitter that splits a dataset 70/30 train and test.