20965361344

Committed 13 Jan 2026 04:59PM UTC coverage: 99.457% (-0.2%) from 99.66%

Build # 20965361344

Build Type

Pull #488

github

Committed by

web-flow

Commit Message

Merge 41c6f440c into c064e80bf

Pull Request Pull Request #488: Add multilinear regression

Run Details

66 of 78 new or added lines in 5 files covered. (84.62%)

7 existing lines in 3 files now uncovered.

8614 of 8661 relevant lines covered (99.46%)

223.31 hits per line

Source File
Press 'n' to go to next uncovered line, 'b' for previous

94.74

/src/Statistics/Regression/Regression.php

<?php

namespace MathPHP\Statistics\Regression;

/**
 * Base class for regressions.
 */
abstract class Regression
{
    /**
     * Array of x and y points: [ [x₁₁, x₁₂, x₁ₖ, y₁], [x₂₁, x₂₂, x₂ₖ, y₂], ... ]
     * @var list<array{float, float, ...<float>}>
     */
    protected $points;

    /**
     * X values of the original points
     * @var list<float>
     */
    protected $xs;

    /**
     * Y values of the original points
     * @var list<float>
     */
    protected $ys;

    /**
     * X row values of the original points
     * @var list<non-empty-list<float>>
     */
    protected $xss;

    /**
     * Number of points
     * @var int
     */
    protected $n;

    /**
     * Number of columns in xss
     * @var int
     */
    protected $k;

    /**
     * Constructor - Prepares the data arrays for regression analysis
     *
     * @param list<array{float|non-empty-list<float>, float}> $points [ [ x₁ | [ x₁₁, x₁₂, x₁ₖ ], y₁ ], [ x₂ | [ x₂₁, x₂₂, x₂ₖ ], y₂ ], ... ]
     */
    public function __construct(array $points)
    {
        $this->points = $points;
        $this->n      = \count($points);
        $this->k      = empty($points) ? 0 : (\is_array($points[0]) ? \count($points[0]) : 1);

        // Get list of x points and y points.
        // This will be fine for linear or polynomial regression, where there is only one x.
        $this->xs = \array_map(function ($point) {
            return \is_array($point[0]) ? $point[0][0] : $point[0];
        }, $points);
        $this->ys = \array_map(function ($point) {
            return $point[1];
        }, $points);
        // For multilinear, the format is a list of x for each point.
        $this->xss = \array_map(function (array $point) {
            return \is_array($point[0]) ? $point[0] : [$point[0]];
        }, $points);
    }

    /**
     * Evaluate the regression equation at x
     *
     * @param float $x
     *
     * @return float
     */
    abstract public function evaluate(float $x): float;

    /**
     * Evaluate the regression equation at x vector
     *
     * @param non-empty-list<float> $vector
     * @return float
     */
    public function evaluateVector(array $vector): float
    {
        return $this->evaluate($vector[0]);
    }

    /**
     * Get points
     *
     * @return array<array{float, float}>
     */
    public function getPoints(): array
    {
        return $this->points;
    }

    /**
     * Get Xs (x values of each point)
     *
     * @return list<float> of x values
     */
    public function getXs(): array
    {
        return $this->xs;
    }

    /**
     * Get Ys (y values of each point)
     *
     * @return list<float> of y values
     */
    public function getYs(): array
    {
        return $this->ys;
    }

    /**
     * Get Xss (x vector of each point)
     *
     * @return list<non-empty-list<float>> of x values
     */
    public function getXss(): array
    {
        return $this->xss;
    }

    /**
     * Get sample size (number of points)
     *
     * @return int
     */
    public function getSampleSize(): int
    {
        return $this->n;
    }

    /**
     * Ŷ (yhat)
     * A list of the predicted values of Y given the regression.
     *
     * @return list<float>
     */
    public function yHat(): array
    {
        return \array_map([$this, 'evaluateVector'], $this->xss);
    }
}

1	<?php
2
3	namespace MathPHP\Statistics\Regression;
4
5	/**
6	* Base class for regressions.
7	*/
8	abstract class Regression
9	{
10	/**
11	* Array of x and y points: [ [x₁₁, x₁₂, x₁ₖ, y₁], [x₂₁, x₂₂, x₂ₖ, y₂], ... ]
12	* @var list<array{float, float, ...<float>}>
13	*/
14	protected $points;
15
16	/**
17	* X values of the original points
18	* @var list<float>
19	*/
20	protected $xs;
21
22	/**
23	* Y values of the original points
24	* @var list<float>
25	*/
26	protected $ys;
27
28	/**
29	* X row values of the original points
30	* @var list<non-empty-list<float>>
31	*/
32	protected $xss;
33
34	/**
35	* Number of points
36	* @var int
37	*/
38	protected $n;
39
40	/**
41	* Number of columns in xss
42	* @var int
43	*/
44	protected $k;
45
46	/**
47	* Constructor - Prepares the data arrays for regression analysis
48	*
49	* @param list<array{float\|non-empty-list<float>, float}> $points [ [ x₁ \| [ x₁₁, x₁₂, x₁ₖ ], y₁ ], [ x₂ \| [ x₂₁, x₂₂, x₂ₖ ], y₂ ], ... ]
50	*/
51	public function __construct(array $points)
52	{
53	$this->points = $points;	124✔
54	$this->n = \count($points);	124✔
55	$this->k = empty($points) ? 0 : (\is_array($points[0]) ? \count($points[0]) : 1);	124✔
56
57	// Get list of x points and y points.
58	// This will be fine for linear or polynomial regression, where there is only one x.
59	$this->xs = \array_map(function ($point) {	124✔
60	return \is_array($point[0]) ? $point[0][0] : $point[0];	123✔
61	}, $points);	124✔
62	$this->ys = \array_map(function ($point) {	124✔
63	return $point[1];	123✔
64	}, $points);	124✔
65	// For multilinear, the format is a list of x for each point.
66	$this->xss = \array_map(function (array $point) {	124✔
67	return \is_array($point[0]) ? $point[0] : [$point[0]];	123✔
68	}, $points);	124✔
69	}
70
71	/**
72	* Evaluate the regression equation at x
73	*
74	* @param float $x
75	*
76	* @return float
77	*/
78	abstract public function evaluate(float $x): float;
79
80	/**
81	* Evaluate the regression equation at x vector
82	*
83	* @param non-empty-list<float> $vector
84	* @return float
85	*/
86	public function evaluateVector(array $vector): float
87	{
88	return $this->evaluate($vector[0]);	20✔
89	}
90
91	/**
92	* Get points
93	*
94	* @return array<array{float, float}>
95	*/
96	public function getPoints(): array
97	{
98	return $this->points;	3✔
99	}
100
101	/**
102	* Get Xs (x values of each point)
103	*
104	* @return list<float> of x values
105	*/
106	public function getXs(): array
107	{
108	return $this->xs;	3✔
109	}
110
111	/**
112	* Get Ys (y values of each point)
113	*
114	* @return list<float> of y values
115	*/
116	public function getYs(): array
117	{
118	return $this->ys;	3✔
119	}
120
121	/**
122	* Get Xss (x vector of each point)
123	*
124	* @return list<non-empty-list<float>> of x values
125	*/
126	public function getXss(): array
127	{
NEW 128	return $this->xss;	×
129	}
130
131	/**
132	* Get sample size (number of points)
133	*
134	* @return int
135	*/
136	public function getSampleSize(): int
137	{
138	return $this->n;	5✔
139	}
140
141	/**
142	* Ŷ (yhat)
143	* A list of the predicted values of Y given the regression.
144	*
145	* @return list<float>
146	*/
147	public function yHat(): array
148	{
149	return \array_map([$this, 'evaluateVector'], $this->xss);	21✔
150	}
151	}

markrogoyski / math-php / 20965361344

Source File Press 'n' to go to next uncovered line, 'b' for previous

Source File
Press 'n' to go to next uncovered line, 'b' for previous