evaluator.h File Reference

#include "kernel/evaluator.h"
#include "kernel/gp/src/dataframe.h"
#include "kernel/gp/src/multi_dataset.h"
#include "kernel/gp/src/oracle.h"
#include <future>
#include <optional>
#include <vector>
#include "kernel/gp/src/evaluator_internal.tcc"
#include "kernel/gp/src/evaluator.tcc"

Go to the source code of this file.

Classes
class	ultra::src::evaluator< D >
class	ultra::src::aggregate_evaluator< P, F, D, A, M >
class	ultra::src::mae_error_functor< P >
class	ultra::src::rmae_error_functor< P >
class	ultra::src::mse_error_functor< P >
class	ultra::src::count_error_functor< P >
class	ultra::src::gaussian_evaluator< P >
class	ultra::src::binary_evaluator< P >

Namespaces
namespace	ultra

Concepts
concept	ultra::src::EvaluationDataset
concept	ultra::src::ExampleEvaluator

Typedefs
template<class D>
using	ultra::src::internal::dataset_example_t = typename dataset_example_impl<D>::type
template<Individual P, class F, class D = multi_dataset<dataframe>>
using	ultra::src::avg_error_evaluator
template<Individual P, class F, class D = multi_dataset<dataframe>>
using	ultra::src::sum_error_evaluator
template<Individual P, class F, class D = multi_dataset<dataframe>>
using	ultra::src::avg_score_evaluator
template<Individual P, class F, class D = multi_dataset<dataframe>>
using	ultra::src::sum_score_evaluator
template<Individual P>
using	ultra::src::mae_evaluator = avg_error_evaluator<P, mae_error_functor<P>>
template<Individual P>
using	ultra::src::rmae_evaluator = avg_error_evaluator<P, rmae_error_functor<P>>
template<Individual P>
using	ultra::src::mse_evaluator = avg_error_evaluator<P, mse_error_functor<P>>
template<Individual P>
using	ultra::src::count_error_evaluator = sum_error_evaluator<P, count_error_functor<P>>

Enumerations
enum class	aggregation_mode { sum , average }
enum class	evaluation_mode { error , score }

Variables
template<class F, class P>
constexpr bool	ultra::src::internal::has_regression_predictor_v
	Convenience variable template for has_regression_predictor.
template<class F>
constexpr bool	ultra::src::parallel_safe_example_evaluator = false
template<Individual P>
constexpr bool	ultra::src::parallel_safe_example_evaluator< mae_error_functor< P > > = true
template<Individual P>
constexpr bool	ultra::src::parallel_safe_example_evaluator< rmae_error_functor< P > > = true
template<Individual P>
constexpr bool	ultra::src::parallel_safe_example_evaluator< mse_error_functor< P > > = true
template<Individual P>
constexpr bool	ultra::src::parallel_safe_example_evaluator< count_error_functor< P > > = true

Detailed Description

Remarks

This file is part of ULTRA.

Copyright

Copyright (C) 2024 EOS di Manlio Morini.

License

This Source Code Form is subject to the terms of the Mozilla Public License, v. 2.0. If a copy of the MPL was not distributed with this file, You can obtain one at http://mozilla.org/MPL/2.0/

Typedef Documentation

avg_error_evaluator

template<Individual P, class F, class D = multi_dataset<dataframe>>

using ultra::src::avg_error_evaluator

Initial value:

 aggregate_evaluator<P, F, D,
                                                aggregation_mode::average,
                                                evaluation_mode::error>

Evaluator minimising the error over a dataset.

Template Parameters

P	individual type
F	error functor type
D	dataset type (defaults to multi_dataset<dataframe>)

This class drives the evolution towards the minimum sum of some sort of error.

Fitness values are normalised so that:

• higher is better;
• values are comparable between full and fast evaluation modes.

See also

mse_evaluator, mae_evaluator, rmae_evaluator, count_error_evaluator

avg_score_evaluator

template<Individual P, class F, class D = multi_dataset<dataframe>>

using ultra::src::avg_score_evaluator

Initial value:

 aggregate_evaluator<P, F, D,
                                                aggregation_mode::average,
                                                evaluation_mode::score>

count_error_evaluator

template<Individual P>

using ultra::src::count_error_evaluator = sum_error_evaluator<P, count_error_functor<P>>

Evaluator based on the number of matches.

See also

count_error_functor

mae_evaluator

template<Individual P>

using ultra::src::mae_evaluator = avg_error_evaluator<P, mae_error_functor<P>>

Evaluator based on the mean absolute error.

See also

mae_error_functor

mse_evaluator

template<Individual P>

using ultra::src::mse_evaluator = avg_error_evaluator<P, mse_error_functor<P>>

Evaluator based on the mean squared error.

See also

mse_error_functor

rmae_evaluator

template<Individual P>

using ultra::src::rmae_evaluator = avg_error_evaluator<P, rmae_error_functor<P>>

Evaluator based on the mean of relative differences.

See also

rmae_error_functor

sum_error_evaluator

template<Individual P, class F, class D = multi_dataset<dataframe>>

using ultra::src::sum_error_evaluator

Initial value:

 aggregate_evaluator<P, F, D,
                                                aggregation_mode::sum,
                                                evaluation_mode::error>

sum_score_evaluator

template<Individual P, class F, class D = multi_dataset<dataframe>>

using ultra::src::sum_score_evaluator

Initial value:

 aggregate_evaluator<P, F, D,
                                                aggregation_mode::sum,
                                                evaluation_mode::score>

Variable Documentation

has_regression_predictor_v

template<class F, class P>

bool ultra::src::internal::has_regression_predictor_v

inlineconstexpr

Initial value:

=
  has_regression_predictor<F, P>::value

Convenience variable template for has_regression_predictor.

parallel_safe_example_evaluator

template<class F>

bool ultra::src::parallel_safe_example_evaluator = false

inlineconstexpr

Opts an example functor into parallel evaluation.

A specialisation set to true promises that each result depends only on the current example and that separate functor instances may run concurrently.