io.jenetics.prog.regression

Interface Error<T>

Type Parameters:

T - the sample type

Functional Interface:

This is a functional interface and can therefore be used as the assignment target for a lambda expression or method reference.

@FunctionalInterface
public interface Error<T>

This function calculates the overall error of a given program tree. The error is calculated from the LossFunction and, if desired, the program Complexity.

final Error<Double> error = Error.of(LossFunction::mse, Complexity.ofNodeCount(50));

Since:

5.0

Version:

5.0

Author:

Franz Wilhelmstötter

See Also:

LossFunction, Complexity

Method Summary

Modifier and Type	Method and Description
double	apply(Tree<? extends Op<T>,?> program, T[] calculated, T[] expected) Calculates the overall error of a given program tree.
static <T> Error<T>	of(LossFunction<T> loss) Creates an error function which only uses the given loss function for calculating the program error
static <T> Error<T>	of(LossFunction<T> loss, Complexity<T> complexity) Creates an error function by combining the given loss function and program complexity.
static <T> Error<T>	of(LossFunction<T> loss, Complexity<T> complexity, DoubleBinaryOperator compose) Creates an error function by combining the given loss function and program complexity.

Method Detail

apply

double apply(Tree<? extends Op<T>,?> program,
             T[] calculated,
             T[] expected)

Calculates the overall error of a given program tree. The error is calculated from the LossFunction and, if desired, the program Complexity.

Parameters:

program - the program tree which calculated the calculated values

calculated - the calculated function values

expected - the expected function values

Returns:

the overall program error

Throws:

NullPointerException - if one of the arguments is null

of

static <T> Error<T> of(LossFunction<T> loss)

Creates an error function which only uses the given loss function for calculating the program error

Type Parameters:

T - the sample type

Parameters:

loss - the loss function to use for calculating the program error

Returns:

an error function which uses the loss function for error calculation

Throws:

NullPointerException - if the given loss function is null

of

static <T> Error<T> of(LossFunction<T> loss,
                       Complexity<T> complexity)

Creates an error function by combining the given loss function and program complexity. The loss function and program complexity is combined in the following way: error = loss + loss*complexity. The complexity function penalizes programs which grows to big.

Type Parameters:

T - the sample type

Parameters:

loss - the loss function

complexity - the program complexity measure

Returns:

a new error function by combining the given loss and complexity function

Throws:

NullPointerException - if one of the functions is null

of

static <T> Error<T> of(LossFunction<T> loss,
                       Complexity<T> complexity,
                       DoubleBinaryOperator compose)

Creates an error function by combining the given loss function and program complexity. The loss function and program complexity is combined in the following way: error = loss + loss*complexity. The complexity function penalizes programs which grows to big.

Type Parameters:

T - the sample type

Parameters:

loss - the loss function

complexity - the program complexity measure

compose - the function which composes the loss and complexity function

Returns:

a new error function by combining the given loss and complexity function

Throws:

NullPointerException - if one of the functions is null