1  #region License Information


2  /* HeuristicLab


3  * Copyright (C) 20022015 Heuristic and Evolutionary Algorithms Laboratory (HEAL)


4  *


5  * This file is part of HeuristicLab.


6  *


7  * HeuristicLab is free software: you can redistribute it and/or modify


8  * it under the terms of the GNU General Public License as published by


9  * the Free Software Foundation, either version 3 of the License, or


10  * (at your option) any later version.


11  *


12  * HeuristicLab is distributed in the hope that it will be useful,


13  * but WITHOUT ANY WARRANTY; without even the implied warranty of


14  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the


15  * GNU General Public License for more details.


16  *


17  * You should have received a copy of the GNU General Public License


18  * along with HeuristicLab. If not, see <http://www.gnu.org/licenses/>.


19  */


20  #endregion


21 


22  using System;


23  using System.Collections.Generic;


24  using System.Diagnostics.Contracts;


25  using System.Linq;


26 


27  namespace HeuristicLab.Common {


28  public static class EnumerableStatisticExtensions {


29  /// <summary>


30  /// Calculates the median element of the enumeration.


31  /// </summary>


32  /// <param name="values"></param>


33  /// <returns></returns>


34  public static double Median(this IEnumerable<double> values) {


35  // See unit tests for comparison with naive implementation


36  return Quantile(values, 0.5);


37  }


38 


39  /// <summary>


40  /// Calculates the alphaquantile element of the enumeration.


41  /// </summary>


42  /// <param name="values"></param>


43  /// <returns></returns>


44  public static double Quantile(this IEnumerable<double> values, double alpha) {


45  // See unit tests for comparison with naive implementation


46  double[] valuesArr = values.ToArray();


47  int n = valuesArr.Length;


48  if (n == 0) throw new InvalidOperationException("Enumeration contains no elements.");


49 


50  // "When N is even, statistics books define the median as the arithmetic mean of the elements k = N/2


51  // and k = N/2 + 1 (that is, N/2 from the bottom and N/2 from the top).


52  // If you accept such pedantry, you must perform two separate selections to find these elements."


53 


54  // return the element at Math.Ceiling (if n*alpha is fractional) or the average of two elements if n*alpha is integer.


55  var pos = n * alpha;


56  Contract.Assert(pos >= 0);


57  Contract.Assert(pos < n);


58  bool isInteger = Math.Round(pos).IsAlmost(pos);


59  if (isInteger) {


60  return 0.5 * (Select((int)pos  1, valuesArr) + Select((int)pos, valuesArr));


61  } else {


62  return Select((int)Math.Ceiling(pos)  1, valuesArr);


63  }


64  }


65 


66  // Numerical Recipes in C++, §8.5 Selecting the Mth Largest, O(n)


67  // Given k in [0..n1] returns an array value from array arr[0..n1] such that k array values are


68  // less than or equal to the one returned. The input array will be rearranged to have this value in


69  // location arr[k], with all smaller elements moved to arr[0..k1] (in arbitrary order) and all


70  // larger elements in arr[k+1..n1] (also in arbitrary order).


71  //


72  // Could be changed to Select<T> where T is IComparable but in this case is significantly slower for double values


73  private static double Select(int k, double[] arr) {


74  Contract.Assert(arr.GetLowerBound(0) == 0);


75  Contract.Assert(k >= 0 && k < arr.Length);


76  int i, ir, j, l, mid, n = arr.Length;


77  double a;


78  l = 0;


79  ir = n  1;


80  for (; ; ) {


81  if (ir <= l + 1) {


82  // Active partition contains 1 or 2 elements.


83  if (ir == l + 1 && arr[ir] < arr[l]) {


84  // if (ir == l + 1 && arr[ir].CompareTo(arr[l]) < 0) {


85  // Case of 2 elements.


86  // SWAP(arr[l], arr[ir]);


87  double temp = arr[l];


88  arr[l] = arr[ir];


89  arr[ir] = temp;


90  }


91  return arr[k];


92  } else {


93  mid = (l + ir) >> 1; // Choose median of left, center, and right elements


94  {


95  // SWAP(arr[mid], arr[l + 1]); // as partitioning element a. Also


96  double temp = arr[mid];


97  arr[mid] = arr[l + 1];


98  arr[l + 1] = temp;


99  }


100 


101  if (arr[l] > arr[ir]) {


102  // if (arr[l].CompareTo(arr[ir]) > 0) { // rearrange so that arr[l] arr[ir] <= arr[l+1],


103  // SWAP(arr[l], arr[ir]); . arr[ir] >= arr[l+1]


104  double temp = arr[l];


105  arr[l] = arr[ir];


106  arr[ir] = temp;


107  }


108 


109  if (arr[l + 1] > arr[ir]) {


110  // if (arr[l + 1].CompareTo(arr[ir]) > 0) {


111  // SWAP(arr[l + 1], arr[ir]);


112  double temp = arr[l + 1];


113  arr[l + 1] = arr[ir];


114  arr[ir] = temp;


115  }


116  if (arr[l] > arr[l + 1]) {


117  //if (arr[l].CompareTo(arr[l + 1]) > 0) {


118  // SWAP(arr[l], arr[l + 1]);


119  double temp = arr[l];


120  arr[l] = arr[l + 1];


121  arr[l + 1] = temp;


122 


123  }


124  i = l + 1; // Initialize pointers for partitioning.


125  j = ir;


126  a = arr[l + 1]; // Partitioning element.


127  for (; ; ) { // Beginning of innermost loop.


128  do i++; while (arr[i] < a /* arr[i].CompareTo(a) < 0 */); // Scan up to find element > a.


129  do j; while (arr[j] > a /* arr[j].CompareTo(a) > 0 */); // Scan down to find element < a.


130  if (j < i) break; // Pointers crossed. Partitioning complete.


131  {


132  // SWAP(arr[i], arr[j]);


133  double temp = arr[i];


134  arr[i] = arr[j];


135  arr[j] = temp;


136  }


137  } // End of innermost loop.


138  arr[l + 1] = arr[j]; // Insert partitioning element.


139  arr[j] = a;


140  if (j >= k) ir = j  1; // Keep active the partition that contains the


141  if (j <= k) l = i; // kth element.


142  }


143  }


144  }


145 


146  /// <summary>


147  /// Calculates the range (max  min) of the enumeration.


148  /// </summary>


149  /// <param name="values"></param>


150  /// <returns></returns>


151  public static double Range(this IEnumerable<double> values) {


152  double min = double.PositiveInfinity;


153  double max = double.NegativeInfinity;


154  int i = 0;


155  foreach (var e in values) {


156  if (min > e) min = e;


157  if (max < e) max = e;


158  i++;


159  }


160  if (i < 1) throw new ArgumentException("The enumerable must contain at least two elements", "values");


161  return max  min;


162  }


163 


164 


165  /// <summary>


166  /// Calculates the sample standard deviation of values.


167  /// </summary>


168  /// <param name="values"></param>


169  /// <returns></returns>


170  public static double StandardDeviation(this IEnumerable<double> values) {


171  return Math.Sqrt(Variance(values));


172  }


173 


174  /// <summary>


175  /// Calculates the population standard deviation of values.


176  /// </summary>


177  /// <param name="values"></param>


178  /// <returns></returns>


179  public static double StandardDeviationPop(this IEnumerable<double> values) {


180  return Math.Sqrt(VariancePop(values));


181  }


182 


183  /// <summary>


184  /// Calculates the sample variance of values. (sum (x  x_mean)² / (n1))


185  /// </summary>


186  /// <param name="values"></param>


187  /// <returns></returns>


188  public static double Variance(this IEnumerable<double> values) {


189  return Variance(values, true);


190  }


191 


192  /// <summary>


193  /// Calculates the population variance of values. (sum (x  x_mean)² / n)


194  /// </summary>


195  /// <param name="values"></param>


196  /// <returns></returns>


197  public static double VariancePop(this IEnumerable<double> values) {


198  return Variance(values, false);


199  }


200 


201  private static double Variance(IEnumerable<double> values, bool sampleVariance) {


202  int m_n = 0;


203  double m_oldM = 0.0;


204  double m_newM = 0.0;


205  double m_oldS = 0.0;


206  double m_newS = 0.0;


207  foreach (double x in values) {


208  m_n++;


209  if (m_n == 1) {


210  m_oldM = m_newM = x;


211  m_oldS = 0.0;


212  } else {


213  m_newM = m_oldM + (x  m_oldM) / m_n;


214  m_newS = m_oldS + (x  m_oldM) * (x  m_newM);


215 


216  // set up for next iteration


217  m_oldM = m_newM;


218  m_oldS = m_newS;


219  }


220  }


221 


222  if (m_n == 0) return double.NaN;


223  if (m_n == 1) return 0.0;


224 


225  if (sampleVariance) return m_newS / (m_n  1);


226  else return m_newS / m_n;


227  }


228 


229  public static IEnumerable<double> LimitToRange(this IEnumerable<double> values, double min, double max) {


230  if (min > max) throw new ArgumentException(string.Format("Minimum {0} is larger than maximum {1}.", min, max));


231  foreach (var x in values) {


232  if (double.IsNaN(x)) yield return (max + min) / 2.0;


233  else if (x < min) yield return min;


234  else if (x > max) yield return max;


235  else yield return x;


236  }


237  }


238  }


239  }

