001
002
003
004
005
006
007
008
009
010
011
012
013
014
015
016
017
018
019
020
021
022
023
024
025
026
027
028
029
030
031
032
033
034
035
036
037
038
039
040
041
042
043
044
045
046
047
048
049
050
051
052
053
054
055
056
057
058
059
060
061
062
063
064
065
066
067
068
069
070
071
072
073
074
075
076
077
078
079
080
081
082
083
084
085
086
087
088
089
090
091
092
093
094
095
096
097
098
099
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
// Exercise 1.1.36
package algs11;
import stdlib.*;

/*

1.1.36 Empirical shuffle check.
Run computational experiments to check that our shuffling code on page 32 works as advertised.
Write a program ShuffleTest that takes commandline arguments M and N, does N shuffles of an
array of size M that is initialized with a[i] = i before each shuffle, and prints an M-by-M table
such that row i gives the number of times i wound up in position j for all j. All entries in the
array should be close to N/M.

Suppose we start with N=3 and run the shuffling test four times,
resulting in the arrays:

  [2, 0, 1]
  [2, 1, 0]
  [2, 0, 1]
  [2, 0, 1]

(This is not very likely, but it is possible)
Then the resulting count array would be:

  0: 0 3 1
  1: 0 1 3
  2: 4 0 0

If we normalize by dividing by the number of times we ran the test,
then we get

  0: 0.000 0.750 0.250
  1: 0.000 0.250 0.750
  2: 1.000 0.000 0.000

That is:
element "2" was always in the first position;
element "0" was 75/25 between second and third position;
and likewise element "1".

You could compute the standard deviation for each of these, which
would be highest for "2".

--------------------------------------------------------------------------------
Here is a more detailed example

initially: counts = {
  {0,0,0},  // position of element 0 in shuffled array
  {0,0,0},  // position of element 1 in shuffled array
  {0,0,0} } // position of element 2 in shuffled array

Iteration 1: test=0
a. initialize array; a={0,1,2}
b. shuffle array; a={1,0,2}
c. increment counts = {
  {0,1,0},
  {1,0,0},
  {0,0,1} }

Iteration 2: test=1
a. initialize array; a={0,1,2}
b. shuffle array; a={0,2,1}
c. increment counts = {
  {1,1,0},
  {1,0,1},
  {0,1,1} }

Iteration 2: test=2
a. initialize array; a={0,1,2}
b. shuffle array; a={0,1,2}
c. increment counts = {
  {2,1,0},
  {1,1,1},
  {0,1,2} }

 */

public class MyShuffleTest {
  public static void main (String[] args) {
    args = new String[] { "3", "40000" };

    int N = Integer.parseInt (args[0]); // size of the array to shuffle
    int numTests = Integer.parseInt (args[1]); // number of times to shuffle

    if (N > Byte.MAX_VALUE) throw new RuntimeException ();

    byte[] a = new byte[N]; // the array
    double[][] counts = new double[N][N]; // a place to put the results

    for (int test = 0; test < numTests; test++) {
      for (byte b = 0; b < a.length; b++)
        a[b] = b;
      shuffle (a);
      if (numTests < 10) StdOut.println (java.util.Arrays.toString (a)); // for debugging: print the shuffled array
      incrementCounts (counts, a);
    }
    printCounts (counts, numTests);
  }

  private static void swap (byte[] a, int i, int j) {
    byte temp = a[i];
    a[i] = a[j];
    a[j] = temp;
  }
  private static void shuffle (byte[] a) {
    int N = a.length;
    for (int i = 0; i < N; i++)
      swap (a, i, i + StdRandom.uniform (N - i));
  }
  private static void incrementCounts (double[][] counts, byte[] a) {
    // count[b][i] is the number of times that byte b is placed in position i
    // TODO
  }
  // This function destroys the counts matrix
  private static void printCounts (double[][] counts, int numTests) {
    // TODO
  }
}