My Chess Playing Computer Programs by James Pate Williams, Jr.

I started creating computer programs the Summer Semester of 2002 at Auburn University. I took a course named “Hand Held Software Development”. The course was taught by my research advisor Associate Professor Richard O. Chapman. I created a distributed chess playing client-server Internet system for human chess players. The program was built using the Palm Pilot’s Palm OS and its C language compiler. Later in circa 2006 I built a rule based and neural network chess program for a computer to play itself and for a human observer (voyeur). The language was Java. Then in 2015 I translated and enhanced my Java program by rewriting the code in C#. Below are pictures of one game.

Word to Words Game Solution by James Pate Williams, Jr.

The description of this alphabetic letter game is very facile. Given a word make as many other words as possible using the letters of the given initial word. But first before we enumerate the game solution, we need to refresh the reader’s memory of some elementary mathematics.
The binary number system also known as the base 2 number system is used by computers to perform arithmetic. The digits in the binary number system are 0 and 1. The numbers 0 to 15 in binary using only four binary digits are where ^ is the exponentiation operator (raising a number to a power) are:
0	0000
1	0001	2 ^ 0 = 1
2	0010	2 ^ 1 = 2
3	0011	2 ^ 1 + 2 ^ 0 = 2 + 1 = 3
4	0100	2 ^ 2 = 4
5	0101	2 ^ 2 + 2 ^ 0 = 4 + 1 = 5
6	0110	2 ^ 2 + 2 ^ 1 = 4 + 2 = 6
7	0111	2 ^ 2 + 2 ^ 1 + 2 ^ 0 = 4 + 2 + 1 = 7
8	1000	2 ^ 3 = 8
9	1001	2 ^ 3 + 2 ^ 0 = 8 + 1 = 9
10	1010	2 ^ 3 + 2 ^ 1 = 8 + 2 = 10
11	1011	2 ^ 3 + 2 ^ 1 + 2 ^ 0 = 8 + 2 + 1 = 11
12	1100	2 ^ 3 + 2 ^ 2 = 8 + 4 = 12
13	1101	2 ^ 3 + 2 ^ 2 + 2 ^ 0 = 8 + 4 + 1 = 13
14	1110	2 ^ 3 + 2 ^ 2 + 2 ^ 1 = 8 + 4 + 2 = 14
15	1111	2 ^ 3 + 2 ^ 2 + 2 ^ 1 + 2 ^ 0 = 8 + 4 + 2 + 1 = 15

An algorithm to convert a base 10 (decimal) number to base 2 (binary) number is given below:
Input n a base 10 number
Output b[0], b[1], b[2], … a finite binary string representing the decimal number

Integer i = 0

Do 
Integer nmod2 = n mod 2
Integer ndiv2 = n / 2

b[i] = nmod2 + ‘0’
i = i + 1

n = ndiv2

While n > 0

The b[i] will be in reverse order. For example, convert 12 from decimal to using four binary digits:
12 mod 2 = 0
12 div 2 = 6

b[0] = ‘0’

i = 1

n = 6

6 mod 2 = 0
6 div 2 = 3

b[1] = ‘0’

i = 2

n = 3

3 mod 2 = 1
3 div 2 = 1

i = 3

b[2] = ‘1’

n = 1

1 mod 2 = 1
1 div 2 = 0

b[3] = ‘1’

n = 0

So, the reversed binary string of digits is “0011”. And after reversing the string we have 12 is represented by the binary digits “1100”.

Next, we need to define a power set and its binary representation. The index power set of 4 objects which has 2 ^ 4 = 16 entries is specified in the following table:
0	0000
1	0001
2	0010
3	0011
4	0100
5	0101
6	0110
7	0111
8	1000
9	1001
10	1010
11	1011
12	1100
13	1101
14	1110
15	1111
A permutation of the set of three indices is given by the following list:
012, 021, 102, 120, 201, 210

A permutation of n objects is a list of n! = n * (n – 1) * … * 2 * 1. A permutation of 4 objects has a list of 24 – 4-digit indices list since 4! = 4 * 3 * 2 * 1 = 24 has the table:
0123, 0132, 0213, 0231, 0312, 0321,
1023, 1032, 1203, 1230, 1320, 1302,
2013, 2031, 2103, 2130, 2301, 2310,
3012, 3021, 3102, 3120, 3201, 3210

Suppose our word is “lost” then we first find the power set:

1	0001	t
2	0010	s
3	0011	st	ts
4	0100	o
5	0101	ot	to
6	0110	os	so
7	0111	ost	ots	sot	sto	tos	tso
8	1000	l
9	1001	lt	tl
10	1010	ls	sl
11	1011	lst	lts	slt	stl	tsl	tls
12	1100	lo	ol
13	1101	lot	lto	olt	otl	tlo	tol	
14	1110	los	lso	slo	sol	osl	ols
15	1111	lost	lots	slot	etc.

Using a dictionary of 152,512 English words my program finds 16 hits for the letters of “lost”:

Dictionary Length: 152512

Word: lost

  0     l
  1     lo
  2     lost
  3     lot
  4     lots
  5     ls
  6     o
  7     s
  8     slot
  9     so
 10     sol
 11     sot
 12     st
 13     t
 14     to
 15     ts

Total letters and/or words 16

Next, we use “tear” as our word:

Dictionary Length: 152512

Word: tear

  0     a
  1     are
  2     art
  3     at
  4     ate
  5     e
  6     ea
  7     ear
  8     eat
  9     era
 10     et
 11     eta
 12     r
 13     rat
 14     rate
 15     re
 16     rt
 17     rte
 18     t
 19     tar
 20     tare
 21     tea
 22     tear
 23     tr

Total letters and/or words 24

Finally, we use the word “company”:

Dictionary Length: 152512

Word: company

  0     a
  1     ac
  2     am
  3     amp
  4     an
  5     any
  6     c
  7     ca
  8     cam
  9     camp
 10     campy
 11     can
 12     canopy
 13     cap
 14     capo
 15     capon
 16     cay
 17     cm
 18     co
 19     com
 20     coma
 21     comp
 22     company
 23     con
 24     cony
 25     cop
 26     copay
 27     copy
 28     coy
 29     cyan
 30     m
 31     ma
 32     mac
 33     man
 34     many
 35     map
 36     may
 37     mayo
 38     mo
 39     moan
 40     mop
 41     mp
 42     my
 43     myna
 44     n
 45     nap
 46     nay
 47     nm
 48     no
 49     o
 50     om
 51     on
 52     op
 53     p
 54     pa
 55     pan
 56     pay
 57     pm
 58     pony
 59     y
 60     ya
 61     yam
 62     yap
 63     yo
 64     yon

Total letters and/or words 65

The C++ program's source code is given below:

// WordToWords.cpp : This file contains the 'main' function. Program execution begins and ends there.
//

#include "pch.h"
#include <algorithm>
#include <fstream>
#include <iomanip>
#include <iostream>
#include <string>
#include <vector>

using namespace std;

vector<string> dictWords;

bool ReadDictionaryFile()
{
    fstream newfile;
 
    newfile.open("C:\\Users\\james\\source\\repos\\WordToWords\\Dictionary.txt", ios::in);

    if (newfile.is_open()) {
        int index = 0, length = 128;
        char cstr[128];

        while (newfile.getline(cstr, length)) {
            string str;

            str.clear();

            for (int i = 0; i < (int)strlen(cstr); i++)
                str.push_back(cstr[i]);

            dictWords.push_back(str);
        }

        newfile.close();

        sort(dictWords.begin(), dictWords.end());
        return true;
    }

    else
        return false;
}

string ConvertBase2(char cstr[], int n, int len)
{
    int count = 0;
    string str, rev;

    do
    {
        int nMod2 = n % 2;
        int nDiv2 = n / 2;

        str.push_back(nMod2 + '0');
        n = nDiv2;
    } while (n > 0);

    n = str.size();

    for (int i = n; i < len; i++)
        str.push_back('0');

    n = str.size();

    for (int i = n - 1; i >= 0; i--)
        if (str[i] == '1')
            rev.push_back(cstr[i]);

    return rev;
}

vector<string> PowerSet(char cstr[], int len)
{
    vector<int> index;
    vector<string> match;

    for (long ps = 0; ps <= pow(2, len); ps++)
    {
        string str = ConvertBase2(cstr, ps, len);
        int psf = 1;

        for (int i = 2; i <= len; i++)
            psf *= i;

        for (int i = 0; i < psf; i++)
        {
            if (binary_search(dictWords.begin(), dictWords.end(), str))
            {
                if (!binary_search(match.begin(), match.end(), str))
                {
                    match.push_back(str);
                    sort(match.begin(), match.end());
                }
            }

            next_permutation(str.begin(), str.end());
        }

        sort(match.begin(), match.end());
    }

    return match;
}

int main()
{
    bool done = false;
    char cstr[128];
    int len;
    string str;
    vector<int> index;
    vector<string> match;

    if (!ReadDictionaryFile())
        return -1;

    cout << "Dictionary Length: " << dictWords.size() << endl << endl;

    cout << "Word: ";
    cin >> cstr;
    cout << endl;

    len = strlen(cstr);

    if (len != 0)
    {
        vector<string> match = PowerSet(cstr, len);

        for (int i = 0; i < match.size(); i++)
        {
            cout << setprecision(3) << setw(3) << i << "\t";
            cout << match[i] << endl;
        }

        cout << endl;
        cout << "Total letters and/or words " << match.size() << endl;
        cout << endl;
    }
}

Schwarzschild Orbit by James Pate Williams, Jr.

The Rosette motion of the perihelion precession of a massive object orbiting a black hole (Schwarzschild solution Einstein’s general relativity field equations is illustrated by the graphs below for varying values of eccentricity of the ellipsoidal orbit 0.00, 0.01, 0.02, 0.03, 0.04, and 0.05. The angular momentum constant is B = 1, and the mass is M = 10.

using System;
using System.Drawing;
using System.Windows.Forms;

namespace SchwarzschildOrbit
{
    public partial class GraphForm : Form
    {
        private double B, B2, B4, M, M2, M3, epsilon;

        public GraphForm(double B, double M, double epsilon)
        {
            InitializeComponent();
            this.B = B;
            this.M = M;
            this.epsilon = epsilon;
            B2 = B * B;
            B4 = B2 * B2;
            M2 = M * M;
            M3 = M * M2;
            panel1.Paint += Panel1_Paint;
        }

        private double u0(double phi)
        {
            return M * (1.0 + epsilon * Math.Cos(phi)) / B2;
        }

        private double u1(double phi)
        {
            return u0(phi) + 3.0 * M3 * (1.0 + epsilon * phi * Math.Sin(phi)
                + epsilon * epsilon * (0.5 - Math.Cos(2.0 * phi) / 6.0)) / B4;
        }

        private double X(double r, double phi)
        {
            return r * Math.Cos(phi);
        }

        private double Y(double r, double phi)
        {
            return r * Math.Sin(phi);
        }

        private void Maximums(out double maxR, out double maxPhi,
            out double XMax, out double XMin, out double YMax, out double YMin)
        {
            double phi = 0.0, r = 0.0, XC = 0.0, YC = 0.0;

            maxPhi = 0.0;
            maxR = double.MinValue;
            XMax = double.MinValue;
            YMax = double.MinValue;
            XMin = double.MaxValue;
            YMin = double.MaxValue;

            while (phi <= 8.0 * Math.PI)
            {
                r = 1.0 / u1(phi);

                if (r > maxR)
                {
                    maxR = r;
                    maxPhi = phi;
                }

                XC = X(r, phi);

                if (XC > XMax)
                    XMax = XC;

                YC = Y(r, phi);

                if (YC > YMax)
                    YMax = YC;

                if (XC < XMin)
                    XMin = XC;

                if (YC < YMin)
                    YMin = YC;

                phi += 0.001;
            }
        }

        private void Minimums(out double minR, out double minPhi,
            out double XMax, out double XMin, out double YMax, out double YMin)
        { 
            double phi = 0.0, r = 0.0, XC = 0.0, YC = 0.0;

            minPhi = 0.0;
            minR = double.MaxValue;
            XMax = double.MinValue;
            YMax = double.MinValue;
            XMin = double.MaxValue;
            YMin = double.MaxValue;

            while (phi <= 8.0 * Math.PI)
            {
                r = 1.0 / u1(phi);

                if (r < minR)
                {
                    minR = r;
                    minPhi = phi;
                }

                XC = X(r, phi);

                if (XC > XMax)
                    XMax = XC;

                YC = Y(r, phi);

                if (YC > YMax)
                    YMax = YC;

                if (XC < XMin)
                    XMin = XC;

                if (YC < YMin)
                    YMin = YC;

                phi += 0.001;
            }
        }

        private void Panel1_Paint(object sender, PaintEventArgs e)
        {
            panel1.Size = ClientSize;
            int width = ClientSize.Width;
            int height = ClientSize.Height;
            int deltaX = width / 6;
            int deltaY = height / 6;
            int minX = deltaX;
            int maxX = 5 * deltaX;
            int minY = deltaY;
            int maxY = 5 * deltaY;
            double maxPhi, minPhi, maxR, minR;
            double XMax, XMin, YMax, YMin;
            double UMax, UMin, VMax, VMin;

            Maximums(out maxR, out maxPhi, out XMax, out XMin, out YMax, out YMin);
            Minimums(out minR, out minPhi, out UMax, out UMin, out VMax, out VMin);
            
            double slopeX = (maxX - minX) / (XMax - XMin);
            double slopeY = (minY - maxY) / (YMax - YMin);
            double interX = minX - slopeX * XMin;
            double interY = maxY - slopeY * YMin;
            double chi = 0.0, eta = 0.0, phi = 0.0, r = 0.0, x, y;
            Graphics g = e.Graphics;
            Pen bp = new Pen(Color.Black);
            SolidBrush rb = new SolidBrush(Color.Red);

            g.Clip = new Region(new Rectangle(minX, minY, maxX - minX + 1, maxY - minY + 1));

            for (int i = 0; i < 5; i++)
                g.DrawLine(bp, (i + 1) * deltaX, minY, (i + 1) * deltaX, maxY);

            for (int i = 0; i < 5; i++)
                g.DrawLine(bp, minX, (i + 1) * deltaY, maxX, (i + 1) * deltaY);

            while (phi <= 8.0 * Math.PI)
            {
                r = 1.0 / u1(phi);
                x = X(r, phi);
                y = Y(r, phi);
                chi = slopeX * x + interX;
                eta = slopeY * y + interY;
                g.FillEllipse(rb, (float)chi, (float)eta, 1, 1);
                phi += 0.001;
            }
        }
    }
}

Word to Word Game by James Pate Williams, Jr.

I implemented a computerized version of a game similar to the board game Boggle. Given a jumble of letters find as many words as possible. Suppose we have the jumbled letters “sldfie” then my C++ program outputs:

Now suppose the scrambled letters are “nifrsmo”:

The program uses permutations and power sets. All permutations of the set of three numbers 1, 2, and 3 are:

123 132 213 231 312 321

The total number of permutations of n objects is n! where n! = n * (n – 1) * … * 2 * 1. So we have 3! = 3 * 2 * 1 = 6 and 4! = 4 * 3 * 2 * 1 = 24.


A power set of three objects can generated by using the binary representation of all binary numbers 0 to 2 * 2 * 2 – 1 = 8 – 1 = 7.

0 = 000, 1 = 001, 2 = 010, 3 = 011, 4 = 100, 5 = 101, 6 = 110, and 7 = 111

My main C++ function in the implementation is as follows:

vector<string> PowerSet(char cstr[], int len)
{
    vector<int> index;
    vector<string> match;

    for (long ps = 0; ps <= pow(2, len); ps++)
    {
        string str = ConvertBase2(cstr, ps, len);
        int psf = 1;

        for (int i = 2; i <= len; i++)
            psf *= i;

        for (int i = 0; i < psf; i++)
        {
            if (binary_search(dictWords.begin(), dictWords.end(), str))
            {
                if (!binary_search(match.begin(), match.end(), str))
                {
                    match.push_back(str);
                    sort(match.begin(), match.end());
                }
            }

            next_permutation(str.begin(), str.end());
        }
    }

    return match;
}

Who Knew? Better Code for Selection Sort

I have been using inferior code for the Selection Sort since 1979. Last night I found the more efficient pseudo code:

Data Structure and Algorithms Selection Sort – Tutorialspoint

Here is my old code for the Selection Sort in C#:

Old Code for the Algorithm

And my new code from the more efficient pseudo code found online:

New and Better Code
Common Swap Function (Procedure)

Both implementations require n * (n – 1) / 2 comparisons which for an array of length 15 is 15 * 14 /2 = 15 * 7 = 105. The second implementation requires typically fewer calls to the swap function.

Selection Sort Test 15-Element Random Array
Selection Sort Test 15-Element Reverse Ordered Array
Selection Sort Test 15-Element Sorted Array

The first number after the unsorted array is the number of comparisons which is always 105 in our 15-element test cases. The second number is the tally of the swap function calls.

Anagrams and Their Computer Solution by James Pate Williams, Jr., BA, BS, MSwE, PhD

An anagram is also known as a word jumble. You take a word and apply a permutation to the word to get an alphabetic jumble of the word. A permutation of three distinct characters is based on three index permutation table:

123 132 213 231 312 321.

So, the scrambling of the word “THE” is as follows:

THE TEH HTE HET ETH EHT.

As you can see there are n-factorial permutations of n objects.

0! = 1

1! = 1

2! = 2 * 1 = 2

3! = 3 * 2 * 1 = 6

4! = 4 * 3 * 2 * 1 = 4 * 3! = 24

Etc. Several years ago I created a program to solve single word anagrams of length less than or equal about a dozen.

12! = 479,001,600

This is about the limit of finding all the permutations of up to length twelve on a desktop computer. The algorithm is extremely easy to understand and implement. First find a suitable list of English words and if the list is unsorted then sort the list alphabetically in ascending order. Hash the dictionary words using a hash table of length 128 * 128 + 128 = 16,512 elements. The dictionary I used has 152,512 words so there are hash table collisions. The hash value is computed using the first three characters of the word in ASCII (7-bit) encoding. Then for each permutation of the anagram a hash value is computed and if the current permutation is found in the hash table the word associated with the hash table entry is returned and the algorithm is finished.

Singleton’s Sorting Algorithm 1979 and 2018 by James Pate Williams, Jr., BS, BS, MSwE, PhD

I first implemented Singleton’s sorting algorithm in the Summer of 1979. The programming language was Data General’s version of Dayton BASIC (Beginner’s All-purpose Symbolic Instruction Code). This variant of BASIC was interpretive like C#, Java, and Pascal. Below is my BASIC version and run-times for double precision numbers in an inverted sequence.

Zoom forward to my current computer and C# programming language in 2018.