Interface DataFrameRegression

All Superinterfaces:

Regression<Tuple>, Serializable, ToDoubleFunction<Tuple>

All Known Implementing Classes:

GradientTreeBoost, LinearModel, RandomForest, RegressionTree

public interface DataFrameRegression
extends Regression<Tuple>

Regression trait on DataFrame.

Nested Class Summary

Nested Classes

Modifier and Type	Interface	Description
static interface	DataFrameRegression.Trainer<M extends DataFrameRegression>	The regression trainer.

Method Summary

Modifier and Type	Method	Description
static DataFrameRegression	ensemble(DataFrameRegression... models)	Return an ensemble of multiple base models to obtain better predictive performance.
Formula	formula()	Returns the model formula.
static DataFrameRegression	of(Formula formula, DataFrame data, Properties params, Regression.Trainer<double[],?> trainer)	Fits a vector regression model on data frame.
default double[]	predict(DataFrame data)	Predicts the dependent variables of a data frame.
StructType	schema()	Returns the schema of predictors.

Methods inherited from interface Regression

applyAsDouble, online, predict, predict, predict, predict, update, update, update

Method Details

formula

Formula formula()

Returns the model formula.

Returns:

the model formula.

schema

StructType schema()

Returns the schema of predictors.

Returns:

the schema of predictors.

predict

default double[] predict(DataFrame data)

Predicts the dependent variables of a data frame.

Parameters:

data - the data frame.

Returns:

the predicted values.

of

static DataFrameRegression of(Formula formula,
 DataFrame data,
 Properties params,
 Regression.Trainer<double[],?> trainer)

Fits a vector regression model on data frame.

Parameters:

formula - a symbolic description of the model to be fitted.

data - the data frame of the explanatory and response variables.

params - the hyperparameters.

trainer - the training lambda.

Returns:

the model.

ensemble

static DataFrameRegression ensemble(DataFrameRegression... models)

Return an ensemble of multiple base models to obtain better predictive performance.

Parameters:

models - the base models.

Returns:

the ensemble model.