Implementation of MS Excel's LOGNORM.DIST(), NORM.S.DIST(), F.DIST(),…

… GAUSS() and GAMMA() functions (#1588) * `GAUSS()` and `GAMMA()`, `NORM.S.DIST()`, `LOGNORM.DIST()` and `F.DIST()` function implementations, and further unit tests for a number of the statistical functions Co-authored-by: Adrien Crivelli <[email protected]>
PHPOffice · Jul 25, 2020 · 57213de · 57213de
1 parent e084e89
commit 57213de
Show file tree

Hide file tree

Showing 25 changed files with 602 additions and 21 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -9,7 +9,7 @@ and this project adheres to [Semantic Versioning](https://semver.org).
 
 ### Added
 
-- Nothing.
+- Implementation of the Excel `LOGNORM.DIST()`, `NORM.S.DIST()`, `GAMMA()` and `GAUSS()` functions. [#1588](https://github.com/PHPOffice/PhpSpreadsheet/pull/1588)
 
 ### Changed
 

diff --git a/src/PhpSpreadsheet/Calculation/Calculation.php b/src/PhpSpreadsheet/Calculation/Calculation.php
@@ -996,7 +996,7 @@ class Calculation
         ],
         'F.DIST' => [
             'category' => Category::CATEGORY_STATISTICAL,
-            'functionCall' => [Functions::class, 'DUMMY'],
+            'functionCall' => [Statistical::class, 'FDIST2'],
             'argumentCount' => '4',
         ],
         'F.DIST.RT' => [
@@ -1133,7 +1133,7 @@ class Calculation
         ],
         'GAMMA' => [
             'category' => Category::CATEGORY_STATISTICAL,
-            'functionCall' => [Functions::class, 'DUMMY'],
+            'functionCall' => [Statistical::class, 'GAMMAFunction'],
             'argumentCount' => '1',
         ],
         'GAMMADIST' => [
@@ -1168,7 +1168,7 @@ class Calculation
         ],
         'GAUSS' => [
             'category' => Category::CATEGORY_STATISTICAL,
-            'functionCall' => [Functions::class, 'DUMMY'],
+            'functionCall' => [Statistical::class, 'GAUSS'],
             'argumentCount' => '1',
         ],
         'GCD' => [
@@ -1577,7 +1577,7 @@ class Calculation
         ],
         'LOGNORM.DIST' => [
             'category' => Category::CATEGORY_STATISTICAL,
-            'functionCall' => [Functions::class, 'DUMMY'],
+            'functionCall' => [Statistical::class, 'LOGNORMDIST2'],
             'argumentCount' => '4',
         ],
         'LOGNORM.INV' => [
@@ -1782,7 +1782,7 @@ class Calculation
         ],
         'NORM.S.DIST' => [
             'category' => Category::CATEGORY_STATISTICAL,
-            'functionCall' => [Functions::class, 'DUMMY'],
+            'functionCall' => [Statistical::class, 'NORMSDIST2'],
             'argumentCount' => '1,2',
         ],
         'NORMSINV' => [
@@ -3131,8 +3131,8 @@ public static function wrapResult($value)
             }
             //    Return strings wrapped in quotes
             return self::FORMULA_STRING_QUOTE . $value . self::FORMULA_STRING_QUOTE;
-        //    Convert numeric errors to NaN error
         } elseif ((is_float($value)) && ((is_nan($value)) || (is_infinite($value)))) {
+            //    Convert numeric errors to NaN error
             return Functions::NAN();
         }
 
@@ -3774,22 +3774,22 @@ private function _parseFormula($formula, ?Cell $pCell = null)
         $pCellParent = ($pCell !== null) ? $pCell->getWorksheet() : null;
 
         $regexpMatchString = '/^(' . self::CALCULATION_REGEXP_FUNCTION .
-                                '|' . self::CALCULATION_REGEXP_CELLREF .
-                                '|' . self::CALCULATION_REGEXP_NUMBER .
-                                '|' . self::CALCULATION_REGEXP_STRING .
-                                '|' . self::CALCULATION_REGEXP_OPENBRACE .
-                                '|' . self::CALCULATION_REGEXP_NAMEDRANGE .
-                                '|' . self::CALCULATION_REGEXP_ERROR .
-                                ')/sui';
+            '|' . self::CALCULATION_REGEXP_CELLREF .
+            '|' . self::CALCULATION_REGEXP_NUMBER .
+            '|' . self::CALCULATION_REGEXP_STRING .
+            '|' . self::CALCULATION_REGEXP_OPENBRACE .
+            '|' . self::CALCULATION_REGEXP_NAMEDRANGE .
+            '|' . self::CALCULATION_REGEXP_ERROR .
+            ')/sui';
 
         //    Start with initialisation
         $index = 0;
         $stack = new Stack();
         $output = [];
         $expectingOperator = false; //    We use this test in syntax-checking the expression to determine when a
-                                                    //        - is a negation or + is a positive operator rather than an operation
+        //        - is a negation or + is a positive operator rather than an operation
         $expectingOperand = false; //    We use this test in syntax-checking the expression to determine whether an operand
-                                                    //        should be null in a function call
+        //        should be null in a function call
 
         // IF branch pruning
         // currently pending storeKey (last item of the storeKeysStack
@@ -4172,7 +4172,7 @@ private function _parseFormula($formula, ?Cell $pCell = null)
                     ((preg_match('/^' . self::CALCULATION_REGEXP_CELLREF . '.*/Ui', substr($formula, $index), $match)) &&
                         ($output[count($output) - 1]['type'] == 'Cell Reference') ||
                         (preg_match('/^' . self::CALCULATION_REGEXP_NAMEDRANGE . '.*/miu', substr($formula, $index), $match)) &&
-                            ($output[count($output) - 1]['type'] == 'Named Range' || $output[count($output) - 1]['type'] == 'Value')
+                        ($output[count($output) - 1]['type'] == 'Named Range' || $output[count($output) - 1]['type'] == 'Value')
                     )) {
                     while ($stack->count() > 0 &&
                         ($o2 = $stack->last()) &&
@@ -4951,7 +4951,7 @@ private function executeNumericBinaryOperation($operand1, $operand2, $operation,
         } else {
             if ((Functions::getCompatibilityMode() != Functions::COMPATIBILITY_OPENOFFICE) &&
                 ((is_string($operand1) && !is_numeric($operand1) && strlen($operand1) > 0) ||
-                 (is_string($operand2) && !is_numeric($operand2) && strlen($operand2) > 0))) {
+                    (is_string($operand2) && !is_numeric($operand2) && strlen($operand2) > 0))) {
                 $result = Functions::VALUE();
             } else {
                 //    If we're dealing with non-matrix operations, execute the necessary operation
@@ -4980,7 +4980,7 @@ private function executeNumericBinaryOperation($operand1, $operand2, $operation,
 
                             return false;
                         }
-                            $result = $operand1 / $operand2;
+                        $result = $operand1 / $operand2;
 
                         break;
                     //    Power

diff --git a/src/PhpSpreadsheet/Calculation/Statistical.php b/src/PhpSpreadsheet/Calculation/Statistical.php
@@ -779,7 +779,7 @@ public static function BETADIST($value, $alpha, $beta, $rMin = 0, $rMax = 1)
     /**
      * BETAINV.
      *
-     * Returns the inverse of the beta distribution.
+     * Returns the inverse of the Beta distribution.
      *
      * @param float $probability Probability at which you want to evaluate the distribution
      * @param float $alpha Parameter to the distribution
@@ -1475,6 +1475,62 @@ public static function EXPONDIST($value, $lambda, $cumulative)
         return Functions::VALUE();
     }
 
+    private static function betaFunction($a, $b)
+    {
+        return (self::gamma($a) * self::gamma($b)) / self::gamma($a + $b);
+    }
+
+    private static function regularizedIncompleteBeta($value, $a, $b)
+    {
+        return self::incompleteBeta($value, $a, $b) / self::betaFunction($a, $b);
+    }
+
+    /**
+     * F.DIST.
+     *
+     *    Returns the F probability distribution.
+     *    You can use this function to determine whether two data sets have different degrees of diversity.
+     *    For example, you can examine the test scores of men and women entering high school, and determine
+     *        if the variability in the females is different from that found in the males.
+     *
+     * @param float $value Value of the function
+     * @param int $u The numerator degrees of freedom
+     * @param int $v The denominator degrees of freedom
+     * @param bool $cumulative If cumulative is TRUE, F.DIST returns the cumulative distribution function;
+     *                         if FALSE, it returns the probability density function.
+     *
+     * @return float|string
+     */
+    public static function FDIST2($value, $u, $v, $cumulative)
+    {
+        $value = Functions::flattenSingleValue($value);
+        $u = Functions::flattenSingleValue($u);
+        $v = Functions::flattenSingleValue($v);
+        $cumulative = Functions::flattenSingleValue($cumulative);
+
+        if (is_numeric($value) && is_numeric($u) && is_numeric($v)) {
+            if ($value < 0 || $u < 1 || $v < 1) {
+                return Functions::NAN();
+            }
+
+            $cumulative = (bool) $cumulative;
+            $u = (int) $u;
+            $v = (int) $v;
+
+            if ($cumulative) {
+                $adjustedValue = ($u * $value) / ($u * $value + $v);
+
+                return self::incompleteBeta($adjustedValue, $u / 2, $v / 2);
+            }
+
+            return (self::gamma(($v + $u) / 2) / (self::gamma($u / 2) * self::gamma($v / 2))) *
+                (($u / $v) ** ($u / 2)) *
+                (($value ** (($u - 2) / 2)) / ((1 + ($u / $v) * $value) ** (($u + $v) / 2)));
+        }
+
+        return Functions::VALUE();
+    }
+
     /**
      * FISHER.
      *
@@ -1556,6 +1612,27 @@ public static function FORECAST($xValue, $yValues, $xValues)
         return $bestFitLinear->getValueOfYForX($xValue);
     }
 
+    /**
+     * GAMMA.
+     *
+     * Return the gamma function value.
+     *
+     * @param float $value
+     *
+     * @return float|string The result, or a string containing an error
+     */
+    public static function GAMMAFunction($value)
+    {
+        $value = Functions::flattenSingleValue($value);
+        if (!is_numeric($value)) {
+            return Functions::VALUE();
+        } elseif ((((int) $value) == ((float) $value)) && $value <= 0.0) {
+            return Functions::NAN();
+        }
+
+        return self::gamma($value);
+    }
+
     /**
      * GAMMADIST.
      *
@@ -1593,7 +1670,7 @@ public static function GAMMADIST($value, $a, $b, $cumulative)
     /**
      * GAMMAINV.
      *
-     * Returns the inverse of the beta distribution.
+     * Returns the inverse of the Gamma distribution.
      *
      * @param float $probability Probability at which you want to evaluate the distribution
      * @param float $alpha Parameter to the distribution
@@ -1677,6 +1754,26 @@ public static function GAMMALN($value)
         return Functions::VALUE();
     }
 
+    /**
+     * GAUSS.
+     *
+     * Calculates the probability that a member of a standard normal population will fall between
+     *     the mean and z standard deviations from the mean.
+     *
+     * @param float $value
+     *
+     * @return float|string The result, or a string containing an error
+     */
+    public static function GAUSS($value)
+    {
+        $value = Functions::flattenSingleValue($value);
+        if (!is_numeric($value)) {
+            return Functions::VALUE();
+        }
+
+        return self::NORMDIST($value, 0, 1, true) - 0.5;
+    }
+
     /**
      * GEOMEAN.
      *
@@ -2117,6 +2214,42 @@ public static function LOGNORMDIST($value, $mean, $stdDev)
         return Functions::VALUE();
     }
 
+    /**
+     * LOGNORM.DIST.
+     *
+     * Returns the lognormal distribution of x, where ln(x) is normally distributed
+     * with parameters mean and standard_dev.
+     *
+     * @param float $value
+     * @param float $mean
+     * @param float $stdDev
+     * @param bool $cumulative
+     *
+     * @return float|string The result, or a string containing an error
+     */
+    public static function LOGNORMDIST2($value, $mean, $stdDev, $cumulative = false)
+    {
+        $value = Functions::flattenSingleValue($value);
+        $mean = Functions::flattenSingleValue($mean);
+        $stdDev = Functions::flattenSingleValue($stdDev);
+        $cumulative = (bool) Functions::flattenSingleValue($cumulative);
+
+        if ((is_numeric($value)) && (is_numeric($mean)) && (is_numeric($stdDev))) {
+            if (($value <= 0) || ($stdDev <= 0)) {
+                return Functions::NAN();
+            }
+
+            if ($cumulative === true) {
+                return self::NORMSDIST2((log($value) - $mean) / $stdDev, true);
+            }
+
+            return (1 / (sqrt(2 * M_PI) * $stdDev * $value)) *
+                exp(0 - ((log($value) - $mean) ** 2 / (2 * $stdDev ** 2)));
+        }
+
+        return Functions::VALUE();
+    }
+
     /**
      * MAX.
      *
@@ -2623,10 +2756,36 @@ public static function NORMINV($probability, $mean, $stdDev)
     public static function NORMSDIST($value)
     {
         $value = Functions::flattenSingleValue($value);
+        if (!is_numeric($value)) {
+            return Functions::VALUE();
+        }
 
         return self::NORMDIST($value, 0, 1, true);
     }
 
+    /**
+     * NORM.S.DIST.
+     *
+     * Returns the standard normal cumulative distribution function. The distribution has
+     * a mean of 0 (zero) and a standard deviation of one. Use this function in place of a
+     * table of standard normal curve areas.
+     *
+     * @param float $value
+     * @param bool $cumulative
+     *
+     * @return float|string The result, or a string containing an error
+     */
+    public static function NORMSDIST2($value, $cumulative)
+    {
+        $value = Functions::flattenSingleValue($value);
+        if (!is_numeric($value)) {
+            return Functions::VALUE();
+        }
+        $cumulative = (bool) Functions::flattenSingleValue($cumulative);
+
+        return self::NORMDIST($value, 0, 1, $cumulative);
+    }
+
     /**
      * NORMSINV.
      *

diff --git a/tests/PhpSpreadsheetTests/Calculation/Functions/Statistical/FDist2Test.php b/tests/PhpSpreadsheetTests/Calculation/Functions/Statistical/FDist2Test.php
@@ -0,0 +1,25 @@
+<?php
+
+namespace PhpOffice\PhpSpreadsheetTests\Calculation\Functions\Statistical;
+
+use PhpOffice\PhpSpreadsheet\Calculation\Statistical;
+use PHPUnit\Framework\TestCase;
+
+class FDist2Test extends TestCase
+{
+    /**
+     * @dataProvider providerFDIST2
+     *
+     * @param mixed $expectedResult
+     */
+    public function testFDIST2($expectedResult, ...$args): void
+    {
+        $result = Statistical::FDIST2(...$args);
+        self::assertEqualsWithDelta($expectedResult, $result, 1E-12);
+    }
+
+    public function providerFDIST2(): array
+    {
+        return require 'tests/data/Calculation/Statistical/FDIST2.php';
+    }
+}
diff --git a/tests/PhpSpreadsheetTests/Calculation/Functions/Statistical/GammaTest.php b/tests/PhpSpreadsheetTests/Calculation/Functions/Statistical/GammaTest.php
@@ -0,0 +1,26 @@
+<?php
+
+namespace PhpOffice\PhpSpreadsheetTests\Calculation\Functions\Statistical;
+
+use PhpOffice\PhpSpreadsheet\Calculation\Statistical;
+use PHPUnit\Framework\TestCase;
+
+class GammaTest extends TestCase
+{
+    /**
+     * @dataProvider providerGAMMA
+     *
+     * @param mixed $expectedResult
+     * @param mixed $testValue
+     */
+    public function testGAMMA($expectedResult, $testValue): void
+    {
+        $result = Statistical::GAMMAFunction($testValue);
+        self::assertEqualsWithDelta($expectedResult, $result, 1E-12);
+    }
+
+    public function providerGAMMA(): array
+    {
+        return require 'tests/data/Calculation/Statistical/GAMMA.php';
+    }
+}