433 lines
10 KiB
PHP
433 lines
10 KiB
PHP
<?php
|
|
/**
|
|
* PHPExcel
|
|
*
|
|
* Copyright (c) 2006 - 2014 PHPExcel
|
|
*
|
|
* This library is free software; you can redistribute it and/or
|
|
* modify it under the terms of the GNU Lesser General Public
|
|
* License as published by the Free Software Foundation; either
|
|
* version 2.1 of the License, or (at your option) any later version.
|
|
*
|
|
* This library is distributed in the hope that it will be useful,
|
|
* but WITHOUT ANY WARRANTY; without even the implied warranty of
|
|
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
|
|
* Lesser General Public License for more details.
|
|
*
|
|
* You should have received a copy of the GNU Lesser General Public
|
|
* License along with this library; if not, write to the Free Software
|
|
* Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
|
|
*
|
|
* @category PHPExcel
|
|
* @package PHPExcel_Shared_Trend
|
|
* @copyright Copyright (c) 2006 - 2014 PHPExcel (http://www.codeplex.com/PHPExcel)
|
|
* @license http://www.gnu.org/licenses/old-licenses/lgpl-2.1.txt LGPL
|
|
* @version ##VERSION##, ##DATE##
|
|
*/
|
|
|
|
|
|
/**
|
|
* PHPExcel_Best_Fit
|
|
*
|
|
* @category PHPExcel
|
|
* @package PHPExcel_Shared_Trend
|
|
* @copyright Copyright (c) 2006 - 2014 PHPExcel (http://www.codeplex.com/PHPExcel)
|
|
*/
|
|
class PHPExcel_Best_Fit
|
|
{
|
|
/**
|
|
* Indicator flag for a calculation error
|
|
*
|
|
* @var boolean
|
|
**/
|
|
protected $_error = False;
|
|
|
|
/**
|
|
* Algorithm type to use for best-fit
|
|
*
|
|
* @var string
|
|
**/
|
|
protected $_bestFitType = 'undetermined';
|
|
|
|
/**
|
|
* Number of entries in the sets of x- and y-value arrays
|
|
*
|
|
* @var int
|
|
**/
|
|
protected $_valueCount = 0;
|
|
|
|
/**
|
|
* X-value dataseries of values
|
|
*
|
|
* @var float[]
|
|
**/
|
|
protected $_xValues = array();
|
|
|
|
/**
|
|
* Y-value dataseries of values
|
|
*
|
|
* @var float[]
|
|
**/
|
|
protected $_yValues = array();
|
|
|
|
/**
|
|
* Flag indicating whether values should be adjusted to Y=0
|
|
*
|
|
* @var boolean
|
|
**/
|
|
protected $_adjustToZero = False;
|
|
|
|
/**
|
|
* Y-value series of best-fit values
|
|
*
|
|
* @var float[]
|
|
**/
|
|
protected $_yBestFitValues = array();
|
|
|
|
protected $_goodnessOfFit = 1;
|
|
|
|
protected $_stdevOfResiduals = 0;
|
|
|
|
protected $_covariance = 0;
|
|
|
|
protected $_correlation = 0;
|
|
|
|
protected $_SSRegression = 0;
|
|
|
|
protected $_SSResiduals = 0;
|
|
|
|
protected $_DFResiduals = 0;
|
|
|
|
protected $_F = 0;
|
|
|
|
protected $_slope = 0;
|
|
|
|
protected $_slopeSE = 0;
|
|
|
|
protected $_intersect = 0;
|
|
|
|
protected $_intersectSE = 0;
|
|
|
|
protected $_Xoffset = 0;
|
|
|
|
protected $_Yoffset = 0;
|
|
|
|
|
|
public function getError() {
|
|
return $this->_error;
|
|
} // function getBestFitType()
|
|
|
|
|
|
public function getBestFitType() {
|
|
return $this->_bestFitType;
|
|
} // function getBestFitType()
|
|
|
|
|
|
/**
|
|
* Return the Y-Value for a specified value of X
|
|
*
|
|
* @param float $xValue X-Value
|
|
* @return float Y-Value
|
|
*/
|
|
public function getValueOfYForX($xValue) {
|
|
return False;
|
|
} // function getValueOfYForX()
|
|
|
|
|
|
/**
|
|
* Return the X-Value for a specified value of Y
|
|
*
|
|
* @param float $yValue Y-Value
|
|
* @return float X-Value
|
|
*/
|
|
public function getValueOfXForY($yValue) {
|
|
return False;
|
|
} // function getValueOfXForY()
|
|
|
|
|
|
/**
|
|
* Return the original set of X-Values
|
|
*
|
|
* @return float[] X-Values
|
|
*/
|
|
public function getXValues() {
|
|
return $this->_xValues;
|
|
} // function getValueOfXForY()
|
|
|
|
|
|
/**
|
|
* Return the Equation of the best-fit line
|
|
*
|
|
* @param int $dp Number of places of decimal precision to display
|
|
* @return string
|
|
*/
|
|
public function getEquation($dp=0) {
|
|
return False;
|
|
} // function getEquation()
|
|
|
|
|
|
/**
|
|
* Return the Slope of the line
|
|
*
|
|
* @param int $dp Number of places of decimal precision to display
|
|
* @return string
|
|
*/
|
|
public function getSlope($dp=0) {
|
|
if ($dp != 0) {
|
|
return round($this->_slope,$dp);
|
|
}
|
|
return $this->_slope;
|
|
} // function getSlope()
|
|
|
|
|
|
/**
|
|
* Return the standard error of the Slope
|
|
*
|
|
* @param int $dp Number of places of decimal precision to display
|
|
* @return string
|
|
*/
|
|
public function getSlopeSE($dp=0) {
|
|
if ($dp != 0) {
|
|
return round($this->_slopeSE,$dp);
|
|
}
|
|
return $this->_slopeSE;
|
|
} // function getSlopeSE()
|
|
|
|
|
|
/**
|
|
* Return the Value of X where it intersects Y = 0
|
|
*
|
|
* @param int $dp Number of places of decimal precision to display
|
|
* @return string
|
|
*/
|
|
public function getIntersect($dp=0) {
|
|
if ($dp != 0) {
|
|
return round($this->_intersect,$dp);
|
|
}
|
|
return $this->_intersect;
|
|
} // function getIntersect()
|
|
|
|
|
|
/**
|
|
* Return the standard error of the Intersect
|
|
*
|
|
* @param int $dp Number of places of decimal precision to display
|
|
* @return string
|
|
*/
|
|
public function getIntersectSE($dp=0) {
|
|
if ($dp != 0) {
|
|
return round($this->_intersectSE,$dp);
|
|
}
|
|
return $this->_intersectSE;
|
|
} // function getIntersectSE()
|
|
|
|
|
|
/**
|
|
* Return the goodness of fit for this regression
|
|
*
|
|
* @param int $dp Number of places of decimal precision to return
|
|
* @return float
|
|
*/
|
|
public function getGoodnessOfFit($dp=0) {
|
|
if ($dp != 0) {
|
|
return round($this->_goodnessOfFit,$dp);
|
|
}
|
|
return $this->_goodnessOfFit;
|
|
} // function getGoodnessOfFit()
|
|
|
|
|
|
public function getGoodnessOfFitPercent($dp=0) {
|
|
if ($dp != 0) {
|
|
return round($this->_goodnessOfFit * 100,$dp);
|
|
}
|
|
return $this->_goodnessOfFit * 100;
|
|
} // function getGoodnessOfFitPercent()
|
|
|
|
|
|
/**
|
|
* Return the standard deviation of the residuals for this regression
|
|
*
|
|
* @param int $dp Number of places of decimal precision to return
|
|
* @return float
|
|
*/
|
|
public function getStdevOfResiduals($dp=0) {
|
|
if ($dp != 0) {
|
|
return round($this->_stdevOfResiduals,$dp);
|
|
}
|
|
return $this->_stdevOfResiduals;
|
|
} // function getStdevOfResiduals()
|
|
|
|
|
|
public function getSSRegression($dp=0) {
|
|
if ($dp != 0) {
|
|
return round($this->_SSRegression,$dp);
|
|
}
|
|
return $this->_SSRegression;
|
|
} // function getSSRegression()
|
|
|
|
|
|
public function getSSResiduals($dp=0) {
|
|
if ($dp != 0) {
|
|
return round($this->_SSResiduals,$dp);
|
|
}
|
|
return $this->_SSResiduals;
|
|
} // function getSSResiduals()
|
|
|
|
|
|
public function getDFResiduals($dp=0) {
|
|
if ($dp != 0) {
|
|
return round($this->_DFResiduals,$dp);
|
|
}
|
|
return $this->_DFResiduals;
|
|
} // function getDFResiduals()
|
|
|
|
|
|
public function getF($dp=0) {
|
|
if ($dp != 0) {
|
|
return round($this->_F,$dp);
|
|
}
|
|
return $this->_F;
|
|
} // function getF()
|
|
|
|
|
|
public function getCovariance($dp=0) {
|
|
if ($dp != 0) {
|
|
return round($this->_covariance,$dp);
|
|
}
|
|
return $this->_covariance;
|
|
} // function getCovariance()
|
|
|
|
|
|
public function getCorrelation($dp=0) {
|
|
if ($dp != 0) {
|
|
return round($this->_correlation,$dp);
|
|
}
|
|
return $this->_correlation;
|
|
} // function getCorrelation()
|
|
|
|
|
|
public function getYBestFitValues() {
|
|
return $this->_yBestFitValues;
|
|
} // function getYBestFitValues()
|
|
|
|
|
|
protected function _calculateGoodnessOfFit($sumX,$sumY,$sumX2,$sumY2,$sumXY,$meanX,$meanY, $const) {
|
|
$SSres = $SScov = $SScor = $SStot = $SSsex = 0.0;
|
|
foreach($this->_xValues as $xKey => $xValue) {
|
|
$bestFitY = $this->_yBestFitValues[$xKey] = $this->getValueOfYForX($xValue);
|
|
|
|
$SSres += ($this->_yValues[$xKey] - $bestFitY) * ($this->_yValues[$xKey] - $bestFitY);
|
|
if ($const) {
|
|
$SStot += ($this->_yValues[$xKey] - $meanY) * ($this->_yValues[$xKey] - $meanY);
|
|
} else {
|
|
$SStot += $this->_yValues[$xKey] * $this->_yValues[$xKey];
|
|
}
|
|
$SScov += ($this->_xValues[$xKey] - $meanX) * ($this->_yValues[$xKey] - $meanY);
|
|
if ($const) {
|
|
$SSsex += ($this->_xValues[$xKey] - $meanX) * ($this->_xValues[$xKey] - $meanX);
|
|
} else {
|
|
$SSsex += $this->_xValues[$xKey] * $this->_xValues[$xKey];
|
|
}
|
|
}
|
|
|
|
$this->_SSResiduals = $SSres;
|
|
$this->_DFResiduals = $this->_valueCount - 1 - $const;
|
|
|
|
if ($this->_DFResiduals == 0.0) {
|
|
$this->_stdevOfResiduals = 0.0;
|
|
} else {
|
|
$this->_stdevOfResiduals = sqrt($SSres / $this->_DFResiduals);
|
|
}
|
|
if (($SStot == 0.0) || ($SSres == $SStot)) {
|
|
$this->_goodnessOfFit = 1;
|
|
} else {
|
|
$this->_goodnessOfFit = 1 - ($SSres / $SStot);
|
|
}
|
|
|
|
$this->_SSRegression = $this->_goodnessOfFit * $SStot;
|
|
$this->_covariance = $SScov / $this->_valueCount;
|
|
$this->_correlation = ($this->_valueCount * $sumXY - $sumX * $sumY) / sqrt(($this->_valueCount * $sumX2 - pow($sumX,2)) * ($this->_valueCount * $sumY2 - pow($sumY,2)));
|
|
$this->_slopeSE = $this->_stdevOfResiduals / sqrt($SSsex);
|
|
$this->_intersectSE = $this->_stdevOfResiduals * sqrt(1 / ($this->_valueCount - ($sumX * $sumX) / $sumX2));
|
|
if ($this->_SSResiduals != 0.0) {
|
|
if ($this->_DFResiduals == 0.0) {
|
|
$this->_F = 0.0;
|
|
} else {
|
|
$this->_F = $this->_SSRegression / ($this->_SSResiduals / $this->_DFResiduals);
|
|
}
|
|
} else {
|
|
if ($this->_DFResiduals == 0.0) {
|
|
$this->_F = 0.0;
|
|
} else {
|
|
$this->_F = $this->_SSRegression / $this->_DFResiduals;
|
|
}
|
|
}
|
|
} // function _calculateGoodnessOfFit()
|
|
|
|
|
|
protected function _leastSquareFit($yValues, $xValues, $const) {
|
|
// calculate sums
|
|
$x_sum = array_sum($xValues);
|
|
$y_sum = array_sum($yValues);
|
|
$meanX = $x_sum / $this->_valueCount;
|
|
$meanY = $y_sum / $this->_valueCount;
|
|
$mBase = $mDivisor = $xx_sum = $xy_sum = $yy_sum = 0.0;
|
|
for($i = 0; $i < $this->_valueCount; ++$i) {
|
|
$xy_sum += $xValues[$i] * $yValues[$i];
|
|
$xx_sum += $xValues[$i] * $xValues[$i];
|
|
$yy_sum += $yValues[$i] * $yValues[$i];
|
|
|
|
if ($const) {
|
|
$mBase += ($xValues[$i] - $meanX) * ($yValues[$i] - $meanY);
|
|
$mDivisor += ($xValues[$i] - $meanX) * ($xValues[$i] - $meanX);
|
|
} else {
|
|
$mBase += $xValues[$i] * $yValues[$i];
|
|
$mDivisor += $xValues[$i] * $xValues[$i];
|
|
}
|
|
}
|
|
|
|
// calculate slope
|
|
// $this->_slope = (($this->_valueCount * $xy_sum) - ($x_sum * $y_sum)) / (($this->_valueCount * $xx_sum) - ($x_sum * $x_sum));
|
|
$this->_slope = $mBase / $mDivisor;
|
|
|
|
// calculate intersect
|
|
// $this->_intersect = ($y_sum - ($this->_slope * $x_sum)) / $this->_valueCount;
|
|
if ($const) {
|
|
$this->_intersect = $meanY - ($this->_slope * $meanX);
|
|
} else {
|
|
$this->_intersect = 0;
|
|
}
|
|
|
|
$this->_calculateGoodnessOfFit($x_sum,$y_sum,$xx_sum,$yy_sum,$xy_sum,$meanX,$meanY,$const);
|
|
} // function _leastSquareFit()
|
|
|
|
|
|
/**
|
|
* Define the regression
|
|
*
|
|
* @param float[] $yValues The set of Y-values for this regression
|
|
* @param float[] $xValues The set of X-values for this regression
|
|
* @param boolean $const
|
|
*/
|
|
function __construct($yValues, $xValues=array(), $const=True) {
|
|
// Calculate number of points
|
|
$nY = count($yValues);
|
|
$nX = count($xValues);
|
|
|
|
// Define X Values if necessary
|
|
if ($nX == 0) {
|
|
$xValues = range(1,$nY);
|
|
$nX = $nY;
|
|
} elseif ($nY != $nX) {
|
|
// Ensure both arrays of points are the same size
|
|
$this->_error = True;
|
|
return False;
|
|
}
|
|
|
|
$this->_valueCount = $nY;
|
|
$this->_xValues = $xValues;
|
|
$this->_yValues = $yValues;
|
|
} // function __construct()
|
|
|
|
} // class bestFit
|