_order; } // function getOrder() public function getValueOfYForX($xValue) { $retVal = $this->getIntersect(); $slope = $this->getSlope(); foreach($slope as $key => $value) { if ($value != 0.0) { $retVal += $value * pow($xValue, $key + 1); } } return $retVal; } // function getValueOfYForX() public function getValueOfXForY($yValue) { return ($yValue - $this->getIntersect()) / $this->getSlope(); } // function getValueOfXForY() public function getEquation($dp=0) { $slope = $this->getSlope($dp); $intersect = $this->getIntersect($dp); $equation = 'Y = '.$intersect; foreach($slope as $key => $value) { if ($value != 0.0) { $equation .= ' + '.$value.' * X'; if ($key > 0) { $equation .= '^'.($key + 1); } } } return $equation; } // function getEquation() public function getSlope($dp=0) { if ($dp != 0) { $coefficients = array(); foreach($this->_slope as $coefficient) { $coefficients[] = round($coefficient,$dp); } return $coefficients; } return $this->_slope; } // function getSlope() public function getCoefficients($dp=0) { return array_merge(array($this->getIntersect($dp)),$this->getSlope($dp)); } // function getCoefficients() private function _polynomial_regression($order, $yValues, $xValues, $const) { // calculate sums $x_sum = array_sum($xValues); $y_sum = array_sum($yValues); $xx_sum = $xy_sum = 0; for($i = 0; $i < $this->_valueCount; ++$i) { $xy_sum += $xValues[$i] * $yValues[$i]; $xx_sum += $xValues[$i] * $xValues[$i]; $yy_sum += $yValues[$i] * $yValues[$i]; } /* * This routine uses logic from the PHP port of polyfit version 0.1 * written by Michael Bommarito and Paul Meagher * * The function fits a polynomial function of order $order through * a series of x-y data points using least squares. * */ for ($i = 0; $i < $this->_valueCount; ++$i) { for ($j = 0; $j <= $order; ++$j) { $A[$i][$j] = pow($xValues[$i], $j); } } for ($i=0; $i < $this->_valueCount; ++$i) { $B[$i] = array($yValues[$i]); } $matrixA = new Matrix($A); $matrixB = new Matrix($B); $C = $matrixA->solve($matrixB); $coefficients = array(); for($i = 0; $i < $C->m; ++$i) { $r = $C->get($i, 0); if (abs($r) <= pow(10, -9)) { $r = 0; } $coefficients[] = $r; } $this->_intersect = array_shift($coefficients); $this->_slope = $coefficients; $this->_calculateGoodnessOfFit($x_sum,$y_sum,$xx_sum,$yy_sum,$xy_sum); foreach($this->_xValues as $xKey => $xValue) { $this->_yBestFitValues[$xKey] = $this->getValueOfYForX($xValue); } } // function _polynomial_regression() function __construct($order, $yValues, $xValues=array(), $const=True) { if (parent::__construct($yValues, $xValues) !== False) { if ($order < $this->_valueCount) { $this->_bestFitType .= '_'.$order; $this->_order = $order; $this->_polynomial_regression($order, $yValues, $xValues, $const); if (($this->getGoodnessOfFit() < 0.0) || ($this->getGoodnessOfFit() > 1.0)) { $this->_error = True; } } else { $this->_error = True; } } } // function __construct() } // class polynomialBestFit