Benchmarks

The benchmarks module provides a comprehensive collection of test problems for evaluating optimization algorithms and machine learning models. It includes classification datasets, optimization test functions, and benchmark suites.

Contents

ML Datasets

Machine learning datasets from the UCI Machine Learning Repository and other sources. These datasets are commonly used for testing classification algorithms.

Reference: Dua, D. and Graff, C. (2019). UCI Machine Learning Repository. Irvine, CA: University of California, School of Information and Computer Science.

Dataset	Description
`IrisDataset`	Famous dataset with iris flower measurements (150 samples, 4 features, 3 classes)
`WineDataset`	Wine recognition dataset from chemical analysis (178 samples, 13 features, 3 classes)
`BreastCancerDataset`	Breast cancer diagnostic dataset (569 samples, 30 features, 2 classes)
`DigitsDataset`	Handwritten digits recognition (5620 samples, 64 features, 10 classes)
`CreditRiskDataset`	Credit risk prediction dataset (3 features, 2 classes)
`UserKnowladgeDataset`	Student knowledge modeling (403 samples, 5 features, 4 classes)
`BanknoteDataset`	Banknote authentication dataset (1372 samples, 4 features, 2 classes)

IrisDataset

class thefittest.benchmarks.IrisDataset

Bases: Dataset

The Iris dataset - one of the most famous datasets in machine learning.

Contains measurements of iris flowers from three different species.

Features (4):

sepal length in cm
sepal width in cm
petal length in cm
petal width in cm

Classes (3):

Iris-setosa (0)
Iris-versicolor (1)
Iris-virginica (2)

Samples: 150 (50 per class)

References

Fisher R. A.. (1988). Iris. UCI Machine Learning Repository. https://doi.org/10.24432/C56C76.

__init__() → None

get_X() → ndarray[Any, dtype[float64]]

Get feature matrix.

Returns:

NDArray[np.float64]: Feature matrix of shape (n_samples, n_features)

get_X_names() → Dict[int, str]

Get feature names.

Returns:

Dict[int, str]: Dictionary mapping feature indices to feature names

get_y() → ndarray[Any, dtype[int64 | float64]]

Get target values.

Returns:

NDArray[Union[np.int64, np.float64]]: Target array of shape (n_samples,)

get_y_names() → Dict[int, str]

Get class/target names.

Returns:

Dict[int, str]: Dictionary mapping class indices to class names

WineDataset

class thefittest.benchmarks.WineDataset

Bases: Dataset

Wine recognition dataset.

Contains results of a chemical analysis of wines grown in the same region in Italy but derived from three different cultivars.

Features (13):

Alcohol
Malic acid
Ash
Alcalinity of ash
Magnesium
Total phenols
Flavanoids
Nonflavanoid phenols
Proanthocyanins
Color intensity
Hue
OD280/OD315 of diluted wines
Proline

Classes (3): class 1 (0), class 2 (1), class 3 (2)

Samples: 178

References

Aeberhard, Stefan and Forina, M.. (1991). Wine. UCI Machine Learning Repository. https://doi.org/10.24432/C5PC7J.

__init__() → None

get_X() → ndarray[Any, dtype[float64]]

Get feature matrix.

Returns:

NDArray[np.float64]: Feature matrix of shape (n_samples, n_features)

get_X_names() → Dict[int, str]

Get feature names.

Returns:

Dict[int, str]: Dictionary mapping feature indices to feature names

get_y() → ndarray[Any, dtype[int64 | float64]]

Get target values.

Returns:

NDArray[Union[np.int64, np.float64]]: Target array of shape (n_samples,)

get_y_names() → Dict[int, str]

Get class/target names.

Returns:

Dict[int, str]: Dictionary mapping class indices to class names

BreastCancerDataset

class thefittest.benchmarks.BreastCancerDataset

Bases: Dataset

Breast Cancer Wisconsin (Diagnostic) dataset.

Contains features computed from digitized images of fine needle aspirate (FNA) of breast masses. They describe characteristics of the cell nuclei present in the image.

Features (30):

Ten real-valued features computed for each cell nucleus:

Mean: radius, texture, perimeter, area, smoothness, compactness, concavity, concave points, symmetry, fractal dimension
Error: radius error, texture error, perimeter error, area error, smoothness error, compactness error, concavity error, concave points error, symmetry error, fractal dimension error
Worst: worst radius, texture, perimeter, area, smoothness, compactness, concavity, concave points, symmetry, fractal dimension

Classes (2): M (malignant, 0), B (benign, 1)

Samples: 569

References

Wolberg, William, Mangasarian, Olvi, Street, Nick, and Street, W.. (1995). Breast Cancer Wisconsin (Diagnostic). UCI Machine Learning Repository. https://doi.org/10.24432/C5DW2B.

__init__() → None

get_X() → ndarray[Any, dtype[float64]]

Get feature matrix.

Returns:

NDArray[np.float64]: Feature matrix of shape (n_samples, n_features)

get_X_names() → Dict[int, str]

Get feature names.

Returns:

Dict[int, str]: Dictionary mapping feature indices to feature names

get_y() → ndarray[Any, dtype[int64 | float64]]

Get target values.

Returns:

NDArray[Union[np.int64, np.float64]]: Target array of shape (n_samples,)

get_y_names() → Dict[int, str]

Get class/target names.

Returns:

Dict[int, str]: Dictionary mapping class indices to class names

DigitsDataset

class thefittest.benchmarks.DigitsDataset

Bases: Dataset

Optical Recognition of Handwritten Digits dataset.

Contains normalized bitmaps of handwritten digits from 0 to 9.

Features (64): 8x8 pixel values (0-16)

Classes (10): Digits 0-9

Samples: 5620

References

Alpaydin, E. and Kaynak, C.. (1998). Optical Recognition of Handwritten Digits. UCI Machine Learning Repository. https://doi.org/10.24432/C50P49.

__init__() → None

get_X() → ndarray[Any, dtype[float64]]

Get feature matrix.

Returns:

NDArray[np.float64]: Feature matrix of shape (n_samples, n_features)

get_X_names() → Dict[int, str]

Get feature names.

Returns:

Dict[int, str]: Dictionary mapping feature indices to feature names

get_y() → ndarray[Any, dtype[int64 | float64]]

Get target values.

Returns:

NDArray[Union[np.int64, np.float64]]: Target array of shape (n_samples,)

get_y_names() → Dict[int, str]

Get class/target names.

Returns:

Dict[int, str]: Dictionary mapping class indices to class names

CreditRiskDataset

class thefittest.benchmarks.CreditRiskDataset

Bases: Dataset

Credit Risk dataset.

For predicting whether a client is a good or bad credit risk based on financial information.

Features (3):

income
age
loan

Classes (2): good client (0), bad client (1)

References

https://www.kaggle.com/datasets/upadorprofzs/credit-risk

__init__() → None

get_X() → ndarray[Any, dtype[float64]]

Get feature matrix.

Returns:

NDArray[np.float64]: Feature matrix of shape (n_samples, n_features)

get_X_names() → Dict[int, str]

Get feature names.

Returns:

Dict[int, str]: Dictionary mapping feature indices to feature names

get_y() → ndarray[Any, dtype[int64 | float64]]

Get target values.

Returns:

NDArray[Union[np.int64, np.float64]]: Target array of shape (n_samples,)

get_y_names() → Dict[int, str]

Get class/target names.

Returns:

Dict[int, str]: Dictionary mapping class indices to class names

UserKnowladgeDataset

class thefittest.benchmarks.UserKnowladgeDataset

Bases: Dataset

User Knowledge Modeling dataset.

Real dataset about students’ knowledge status about the subject of Electrical DC Machines.

Features (5):

STG: The degree of study time for goal object materials
SCG: The degree of repetition number of user for goal object materials
STR: The degree of study time of user for related objects with goal object
LPR: The exam performance of user for related objects with goal object
PEG: The exam performance of user for goal objects

Classes (4): Very Low (0), Low (1), Middle (2), High (3)

Samples: 403

References

Kahraman, Hamdi, Colak, Ilhami, and Sagiroglu, Seref. (2013). User Knowledge Modeling. UCI Machine Learning Repository. https://doi.org/10.24432/C5231X.

__init__() → None

get_X() → ndarray[Any, dtype[float64]]

Get feature matrix.

Returns:

NDArray[np.float64]: Feature matrix of shape (n_samples, n_features)

get_X_names() → Dict[int, str]

Get feature names.

Returns:

Dict[int, str]: Dictionary mapping feature indices to feature names

get_y() → ndarray[Any, dtype[int64 | float64]]

Get target values.

Returns:

NDArray[Union[np.int64, np.float64]]: Target array of shape (n_samples,)

get_y_names() → Dict[int, str]

Get class/target names.

Returns:

Dict[int, str]: Dictionary mapping class indices to class names

BanknoteDataset

class thefittest.benchmarks.BanknoteDataset

Bases: Dataset

Banknote authentication dataset.

Data extracted from images taken for the evaluation of an authentication procedure for banknotes.

Features (4):

variance of Wavelet Transformed image (continuous)
skewness of Wavelet Transformed image (continuous)
curtosis of Wavelet Transformed image (continuous)
entropy of image (continuous)

Classes (2): not original (0), original (1)

Samples: 1372

References

Lohweg, Volker. (2013). banknote authentication. UCI Machine Learning Repository. https://doi.org/10.24432/C55P57.

__init__() → None

get_X() → ndarray[Any, dtype[float64]]

Get feature matrix.

Returns:

NDArray[np.float64]: Feature matrix of shape (n_samples, n_features)

get_X_names() → Dict[int, str]

Get feature names.

Returns:

Dict[int, str]: Dictionary mapping feature indices to feature names

get_y() → ndarray[Any, dtype[int64 | float64]]

Get target values.

Returns:

NDArray[Union[np.int64, np.float64]]: Target array of shape (n_samples,)

get_y_names() → Dict[int, str]

Get class/target names.

Returns:

Dict[int, str]: Dictionary mapping class indices to class names

Optimization Functions

Classic benchmark functions for testing continuous optimization algorithms. Each function has different characteristics (unimodal/multimodal, separable/non-separable) that challenge different aspects of optimization algorithms.

Function	Description
`Sphere`	Simple quadratic function, unimodal and convex
`Rosenbrock`	Valley-shaped function, unimodal with narrow parabolic valley
`Rastrigin`	Highly multimodal with many local minima
`Ackley`	Multimodal with nearly flat outer region and large hole at center
`Griewank`	Multimodal with many widespread local minima
`Weierstrass`	Continuous nowhere differentiable function with fractal structure
`Schwefe1_2`	Unimodal with non-separable variables
`HighConditionedElliptic`	Unimodal with high condition number
`OneMax`	Simple sum of all variables

Sphere

class thefittest.benchmarks.Sphere

Bases: TestFunction

Sphere function - simple quadratic function.

One of the simplest optimization test functions. It is continuous, convex, and unimodal. The global minimum is at x = [0, 0, …, 0] with f(x) = 0.

Formula: f(x) = sum(x_i^2)

Parameters:

xNDArray[np.float64]: Input array of shape (n_samples, n_dimensions)

Returns:

NDArray[np.float64]: Function values of shape (n_samples,)

f(x: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

build_grid(x: ndarray[Any, dtype[float64]], y: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

Rosenbrock

class thefittest.benchmarks.Rosenbrock

Bases: TestFunction

Rosenbrock function (De Jong’s function 2, Valley function).

A classic optimization test function with a narrow, parabolic valley. The global minimum is inside a long, narrow, parabolic shaped flat valley. Finding the valley is trivial, but convergence to the global minimum is difficult. The global minimum is at x = [1, 1, …, 1] with f(x) = 0.

Formula: f(x) = sum_{i=1}^{n-1} [100(x_{i+1} - x_i^2)^2 + (x_i - 1)^2]

Parameters:

xNDArray[np.float64]: Input array of shape (n_samples, n_dimensions)

Returns:

NDArray[np.float64]: Function values of shape (n_samples,)

f(x: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

build_grid(x: ndarray[Any, dtype[float64]], y: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

Rastrigin

class thefittest.benchmarks.Rastrigin

Bases: TestFunction

Rastrigin function.

A highly multimodal function with a large number of local minima. The function is based on the Sphere function with added cosine modulation to create the local minima. The global minimum is at x = [0, 0, …, 0] with f(x) = 0.

Formula: f(x) = 10n + sum_{i=1}^n [x_i^2 - 10*cos(2*pi*x_i)]

Parameters:

xNDArray[np.float64]: Input array of shape (n_samples, n_dimensions)

Returns:

NDArray[np.float64]: Function values of shape (n_samples,)

f(x: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

build_grid(x: ndarray[Any, dtype[float64]], y: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

Ackley

class thefittest.benchmarks.Ackley

Bases: TestFunction

Ackley function.

A multimodal function with many local minima and a single global minimum. It is characterized by a nearly flat outer region and a large hole at the center. The global minimum is at x = [0, 0, …, 0] with f(x) = 0.

Formula: f(x) = -a*exp(-b*sqrt(sum(x_i^2)/n)) - exp(sum(cos(c*x_i))/n) + a + e where a=20, b=0.2, c=2*pi

Parameters:

xNDArray[np.float64]: Input array of shape (n_samples, n_dimensions)

Returns:

NDArray[np.float64]: Function values of shape (n_samples,)

f(x: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

build_grid(x: ndarray[Any, dtype[float64]], y: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

Griewank

class thefittest.benchmarks.Griewank

Bases: TestFunction

Griewank function.

A multimodal function with many widespread local minima. The number and positioning of local minima is space dependent. The global minimum is at x = [0, 0, …, 0] with f(x) = 0.

Formula: f(x) = 1 + sum_{i=1}^n x_i^2/4000 - prod_{i=1}^n cos(x_i/sqrt(i))

Parameters:

xNDArray[np.float64]: Input array of shape (n_samples, n_dimensions)

Returns:

NDArray[np.float64]: Function values of shape (n_samples,)

f(x: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

build_grid(x: ndarray[Any, dtype[float64]], y: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

Weierstrass

class thefittest.benchmarks.Weierstrass

Bases: TestFunction

Weierstrass function.

A continuous but nowhere differentiable function with a fractal structure. It is highly multimodal with many local optima. The global minimum is at x = [0, 0, …, 0].

Formula: f(x) = sum_{i=1}^n sum_{k=0}^{k_max} [a^k * cos(2*pi*b^k*(x_i+0.5))]

n*sum_{k=0}^{k_max} [a^k * cos(2*pi*b^k*0.5)]

where a=0.5, b=3, k_max=20

Parameters:

xNDArray[np.float64]: Input array of shape (n_samples, n_dimensions)

Returns:

NDArray[np.float64]: Function values of shape (n_samples,)

f(x: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

build_grid(x: ndarray[Any, dtype[float64]], y: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

Schwefe1_2

class thefittest.benchmarks.Schwefe1_2

Bases: TestFunction

Schwefel’s Problem 1.2.

A unimodal function with a single global minimum. The variables are not separable, which makes it harder to optimize than the Sphere function. The global minimum is at x = [0, 0, …, 0] with f(x) = 0.

Formula: f(x) = sum_{i=1}^n (sum_{j=1}^i x_j)^2

Parameters:

xNDArray[np.float64]: Input array of shape (n_samples, n_dimensions)

Returns:

NDArray[np.float64]: Function values of shape (n_samples,)

f(x: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

build_grid(x: ndarray[Any, dtype[float64]], y: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

HighConditionedElliptic

class thefittest.benchmarks.HighConditionedElliptic

Bases: TestFunction

High Conditioned Elliptic function.

A unimodal function with high condition number, making it difficult for optimization algorithms that are sensitive to the scaling of variables. The global minimum is at x = [0, 0, …, 0] with f(x) = 0.

Formula: f(x) = sum_{i=1}^n (10^6)^((i-1)/(n-1)) * x_i^2

Parameters:

xNDArray[np.float64]: Input array of shape (n_samples, n_dimensions)

Returns:

NDArray[np.float64]: Function values of shape (n_samples,)

f(x: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

build_grid(x: ndarray[Any, dtype[float64]], y: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

OneMax

class thefittest.benchmarks.OneMax

Bases: TestFunction

OneMax function - simple sum of all variables.

A basic test function that simply sums all input variables. The global minimum is at x = [0, 0, …, 0] with f(x) = 0.

Parameters:

xNDArray[np.float64]: Input array of shape (n_samples, n_dimensions)

Returns:

NDArray[np.float64]: Function values of shape (n_samples,)

f(x: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

build_grid(x: ndarray[Any, dtype[float64]], y: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

Benchmark Suites

Comprehensive benchmark suites with multiple test functions for systematic algorithm evaluation.

CEC2005

The CEC 2005 Special Session on Real-Parameter Optimization provides 25 test functions organized into categories: unimodal (F1-F5), basic multimodal (F6-F12), expanded (F13-F14), and hybrid composition functions (F15-F25).

Reference: Suganthan, P. N., Hansen, N., Liang, J. J., Deb, K., Chen, Y. P., Auger, A., & Tiwari, S. (2005). Problem Definitions and Evaluation Criteria for the CEC 2005 Special Session on Real-Parameter Optimization.

Module: thefittest.benchmarks.CEC2005

Usage Example:

from thefittest.benchmarks import CEC2005

# Access problem dictionary
problems = CEC2005.problems_dict

# Get F1 (Shifted Sphere)
f1_config = problems["F1"]
function = f1_config["function"]()
bounds = f1_config["bounds"]
optimum = f1_config["optimum"]

Symbolic Regression

A collection of 17 test functions for symbolic regression and genetic programming benchmarks. Functions range from 1D to 2D with varying complexity.

Module: thefittest.benchmarks.symbolicregression17

thefittest.benchmarks.symbolicregression17.z(x: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

Helper function used in F13 and F14.

Combination of three inverse squared terms with different parameters.

Parameters:

xNDArray[np.float64]: Input array

Returns:

NDArray[np.float64]: Function values

thefittest.benchmarks.symbolicregression17.F1(x: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

Complex oscillatory function with exponential decay (1D).

Domain: [-1, 1] Variables: 1

Parameters:

xNDArray[np.float64]: Input array of shape (n_samples, 1)

Returns:

NDArray[np.float64]: Function values of shape (n_samples,)

thefittest.benchmarks.symbolicregression17.F2(x: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

Multi-frequency cosine composition (1D).

Domain: [-1, 1] Variables: 1

Parameters:

xNDArray[np.float64]: Input array of shape (n_samples, 1)

Returns:

NDArray[np.float64]: Function values of shape (n_samples,)

thefittest.benchmarks.symbolicregression17.F3(x: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

Quadratic with cosine terms (2D).

Domain: [-16, 16] Variables: 2

Parameters:

xNDArray[np.float64]: Input array of shape (n_samples, 2)

Returns:

NDArray[np.float64]: Function values of shape (n_samples,)

thefittest.benchmarks.symbolicregression17.F4(x: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

Scaled version of F3 (2D).

Domain: [-16, 16] Variables: 2

Parameters:

xNDArray[np.float64]: Input array of shape (n_samples, 2)

Returns:

NDArray[np.float64]: Function values of shape (n_samples,)

thefittest.benchmarks.symbolicregression17.F5(x: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

Rosenbrock function (2D).

Classic optimization benchmark with a narrow parabolic valley.

Domain: [-2, 2] Variables: 2

Parameters:

xNDArray[np.float64]: Input array of shape (n_samples, 2)

Returns:

NDArray[np.float64]: Function values of shape (n_samples,)

thefittest.benchmarks.symbolicregression17.F6(x: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

Modified Griewank-like function (2D).

Domain: [-16, 16] Variables: 2

Parameters:

xNDArray[np.float64]: Input array of shape (n_samples, 2)

Returns:

NDArray[np.float64]: Function values of shape (n_samples,)

thefittest.benchmarks.symbolicregression17.F7(x: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

Inverse Rosenbrock (2D).

Domain: [-5, 5] Variables: 2

Parameters:

xNDArray[np.float64]: Input array of shape (n_samples, 2)

Returns:

NDArray[np.float64]: Function values of shape (n_samples,)

thefittest.benchmarks.symbolicregression17.F8(x: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

Modified Schaffer function (2D).

Domain: [-10, 10] Variables: 2

Parameters:

xNDArray[np.float64]: Input array of shape (n_samples, 2)

Returns:

NDArray[np.float64]: Function values of shape (n_samples,)

thefittest.benchmarks.symbolicregression17.F9(x: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

Quadratic with multi-frequency cosines (2D).

Domain: [-2.5, 2.5] Variables: 2

Parameters:

xNDArray[np.float64]: Input array of shape (n_samples, 2)

Returns:

NDArray[np.float64]: Function values of shape (n_samples,)

thefittest.benchmarks.symbolicregression17.F10(x: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

Similar to F9 with different domain (2D).

Domain: [-5, 5] Variables: 2

Parameters:

xNDArray[np.float64]: Input array of shape (n_samples, 2)

Returns:

NDArray[np.float64]: Function values of shape (n_samples,)

thefittest.benchmarks.symbolicregression17.F11(x: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

Absolute sine products with inverse term (2D).

Domain: [-4, 4] Variables: 2

Parameters:

xNDArray[np.float64]: Input array of shape (n_samples, 2)

Returns:

NDArray[np.float64]: Function values of shape (n_samples,)

thefittest.benchmarks.symbolicregression17.F12(x: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

Cross-term interaction function (2D).

Domain: [0, 4] Variables: 2

Parameters:

xNDArray[np.float64]: Input array of shape (n_samples, 2)

Returns:

NDArray[np.float64]: Function values of shape (n_samples,)

thefittest.benchmarks.symbolicregression17.F13(x: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

Product of z-functions (2D).

Domain: [0, 4] Variables: 2

Parameters:

xNDArray[np.float64]: Input array of shape (n_samples, 2)

Returns:

NDArray[np.float64]: Function values of shape (n_samples,)

thefittest.benchmarks.symbolicregression17.F14(x: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

Sum of z-functions (2D).

Domain: [0, 4] Variables: 2

Parameters:

xNDArray[np.float64]: Input array of shape (n_samples, 2)

Returns:

NDArray[np.float64]: Function values of shape (n_samples,)

thefittest.benchmarks.symbolicregression17.F15(x: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

Simple quadratic (2D).

Domain: [-5, 5] Variables: 2

Parameters:

xNDArray[np.float64]: Input array of shape (n_samples, 2)

Returns:

NDArray[np.float64]: Function values of shape (n_samples,)

thefittest.benchmarks.symbolicregression17.F16(x: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

Sine with quadratic term (1D).

Domain: [-5, 5] Variables: 1

Parameters:

xNDArray[np.float64]: Input array of shape (n_samples, 1)

Returns:

NDArray[np.float64]: Function values of shape (n_samples,)

thefittest.benchmarks.symbolicregression17.F17(x: ndarray[Any, dtype[float64]]) → ndarray[Any, dtype[float64]]

Linear sine function (1D).

Domain: [-5, 5] Variables: 1

Parameters:

xNDArray[np.float64]: Input array of shape (n_samples, 1)

Returns:

NDArray[np.float64]: Function values of shape (n_samples,)

Usage Example:

from thefittest.benchmarks import symbolicregression17
import numpy as np

# Get F5 (Rosenbrock)
f5_config = symbolicregression17.problems_dict["F5"]
function = f5_config["function"]

# Generate data
X = np.random.uniform(*f5_config["bounds"], size=(100, 2))
y = function(X)