// This is the code that was used to generate the performance
// graphs in the codeproject article.
static void Main(string[] args)
{
Console.WriteLine("Start Tests");
PerformanceTimer pt = new PerformanceTimer();
Random random = new Random();
TBinarySTree bt;
Hashtable ht;
int[] dataSizeArray = new int[22] {
1000, 5000, 10000, 20000, 30000,
40000, 50000, 60000, 80000,
100000, 120000, 140000, 160000, 180000,
200000, 250000, 300000, 400000, 500000,
750000, 1000000, 1000000
};
const int numberOfIntervals = 22;
string[] values = new string[10000000]; // Max number, ever
TextWriter tw = new StreamWriter(@"D:\NET\eLearning\BinaryTree\test.txt");
try {
// Number of trials in a particular interval run
int trials = 20000;
tw.WriteLine("Start Tests, number of intervals: {0}\n", numberOfIntervals);
tw.WriteLine("Number of trials in each interval: {0}\n", trials);
for (int nn = 0; nn < numberOfIntervals; nn++)
{
Console.WriteLine("\nNumber of symbols to insert {0}", dataSizeArray[nn]);
// Generate the keys that will stored in the tables
for (int i = 0; i < dataSizeArray[nn]; i++)
{
values[i] = randomString(random, 10, true);
}
double meanBt = 0; // Mean time for binary search tree
double meanHt = 0; // Mean time for hash table
int index;
double[] timeBt = new double[trials];
double[] timeHt = new double[trials];
// Insert data into binary search tree
bt = new TBinarySTree();
for (int i = 0; i < dataSizeArray[nn]; i++)
{
bt.insert(values[i], i);
}
// Insert data into hash table
ht = new Hashtable();
for (int i = 0; i < dataSizeArray[nn]; i++)
{
ht.Add(values[i], i.ToString());
}
// Data in place, ready to time retrieval.
// Retrieve data from a tree/table 'trial times' Eg
// retrieve 20,000 times (trials) from a
// binary tree that contains 500,000 items (dataSizeArray)
for (int k = 0; k < trials; k++)
{
if (k % 4000 == 0)
{
Console.WriteLine("{0}", k.ToString());
}
// Pick a value at random from the value array, 0
// to the number of values in the dataSizeArray array.
// Eg, assume nn = 5th trial, dataSizeArray[4] = 30000
// We pick a number from a random location in values[0->30000]
// and time how long to takes to retrieve it.
// For each interval we store all the trial times, then
// comput their average and standard deviation.
// Binary Search Tree
index = random.Next(dataSizeArray[nn]);
pt.Start();
bt.findSymbol(values[index]);
pt.Stop();
timeBt[k] = pt.DurationSeconds;
// Hash Table
index = random.Next(dataSizeArray[nn]);
pt.Start();
ht[values[index]].ToString();
pt.Stop();
timeHt[k] = pt.DurationSeconds;
}
// Compute the mean time
for (int i = 0; i < trials; i++)
{
meanBt = meanBt + timeBt[i];
meanHt = meanHt + timeHt[i];
}
meanBt = meanBt / trials;
meanHt = meanHt / trials;
Console.WriteLine();
Console.WriteLine("\nAverage Time for Binary Tree = {0}", meanBt);
// Compute standard deviation
double sd = 0;
for (int i = 0; i < trials; i++)
{
sd = sd + (timeBt[i] - meanBt) * (timeBt[i] - meanBt);
}
sd = Math.Sqrt(sd / trials);
// CV = coefficient of variation
Console.WriteLine("Standard deviation = {0}, CV = {1}", sd, sd / meanBt);
Console.WriteLine("\nAverage time for Hash Table = {0}", meanHt);
sd = 0;
for (int i = 0; i < trials; i++)
{
sd = sd + (timeHt[i] - meanHt) * (timeHt[i] - meanHt);
}
sd = Math.Sqrt(sd / trials);
Console.WriteLine("Standard deviation = {0}, CV = {1}", sd, sd / meanHt);
tw.WriteLine("{0} {1} {2}", dataSizeArray[nn], meanBt, meanHt);
}
} finally {
tw.Close();
}
// Deletion tests
Console.WriteLine("Test Deletion method\n");
bt = new TBinarySTree();
bt.insert("50", 50);
bt.insert("60", 60);
bt.insert("40", 40);
bt.insert("30", 30);
bt.insert("20", 20);
bt.insert("35", 35);
bt.insert("45", 45);
bt.insert("44", 44);
bt.insert("46", 46);
Console.WriteLine("Number of nodes in the tree = {0}\n", bt.count());
Console.WriteLine("Original: " + bt.drawTree());
bt.delete("40");
Console.WriteLine("Delete node 40: " + bt.drawTree());
bt.delete("45");
Console.WriteLine("Delete node 45: " + bt.drawTree());
Console.WriteLine("\nSimple one layered tree");
bt = new TBinarySTree();
bt.insert("50", 50);
bt.insert("20", 20);
bt.insert("90", 90);
Console.WriteLine("\nOriginal: " + bt.drawTree());
bt.delete("50");
Console.WriteLine("Delete node 50: " + bt.drawTree());
bt = new TBinarySTree();
bt.insert("50", 50);
bt.insert("20", 20);
bt.insert("90", 90);
Console.WriteLine("\nOriginal: " + bt.drawTree());
bt.delete("20");
Console.WriteLine("Delete node 20: " + bt.drawTree());
bt = new TBinarySTree();
bt.insert("50", 50);
bt.insert("20", 20);
bt.insert("90", 90);
Console.WriteLine("\nOriginal: " + bt.drawTree());
bt.delete("90");
Console.WriteLine("Delete node 90: " + bt.drawTree());
bt.delete("20");
Console.WriteLine("Delete node 20: " + bt.drawTree());
bt.delete("50");
Console.WriteLine("Delete node 50: " + bt.drawTree());
Console.WriteLine("\n");
bt = new TBinarySTree();
bt.insert("L", 1);
bt.insert("D", 2);
bt.insert("C", 3);
bt.insert("A", 4);
bt.insert("H", 5);
bt.insert("F", 6);
bt.insert("J", 7);
bt.insert("P", 8);
Console.WriteLine("Original: " + bt.drawTree());
bt.delete("J");
Console.WriteLine("Delete J: " + bt.drawTree());
bt.delete("C");
Console.WriteLine("Delete C: " + bt.drawTree());
bt.delete("L");
Console.WriteLine("Delete L: " + bt.drawTree());
bt.delete("D");
Console.WriteLine("Delete D: " + bt.drawTree());
bt.delete("A");
Console.WriteLine("Delete A: " + bt.drawTree());
Console.ReadLine();
}
// This is the code that was used to generate the performance
// graphs in the codeproject article.
static void Main(string[] args) {
Console.WriteLine ("Start Tests");
PerformanceTimer pt = new PerformanceTimer();
Random random = new Random();
TBinarySTree bt;
Hashtable ht;
int[] dataSizeArray = new int[22] { 1000, 5000, 10000, 20000, 30000,
40000, 50000, 60000, 80000,
100000, 120000, 140000, 160000, 180000,
200000, 250000, 300000, 400000, 500000,
750000, 1000000, 1000000};
const int numberOfIntervals = 22;
string[] values = new string[10000000]; // Max number, ever
TextWriter tw = new StreamWriter("c:\\results.txt");
try {
// Number of trials in a particular interval run
int trials = 20000;
tw.WriteLine ("Start Tests, number of intervals: {0}\n", numberOfIntervals);
tw.WriteLine ("Number of trials in each interval: {0}\n", trials);
for (int nn=0; nn<numberOfIntervals; nn++) {
Console.WriteLine ("\nNumber of symbols to insert {0}", dataSizeArray[nn]);
// Generate the keys that will stored in the tables
for (int i=0; i<dataSizeArray[nn]; i++)
values[i] = randomString (random, 10, true);
double meanBt = 0; // Mean time for binary search tree
double meanHt = 0; // Mean time for hash table
int index;
double[] timeBt = new double[trials];
double[] timeHt = new double[trials];
// Insert data into binary search tree
bt = new TBinarySTree();
for (int i=0; i<dataSizeArray[nn]; i++)
bt.insert (values[i], i);
// Insert data into hash table
ht = new Hashtable();
for (int i=0; i<dataSizeArray[nn]; i++)
ht.Add(values[i], i.ToString());
// Data in place, ready to time retrieval.
// Retrieve data from a tree/table 'trial times' Eg
// retrieve 20,000 times (trials) from a
// binary tree that contains 500,000 items (dataSizeArray)
for (int k=0; k<trials; k++) {
if (k % 4000 == 0) Console.WriteLine ("{0}", k.ToString());
// Pick a value at random from the value array, 0
// to the number of values in the dataSizeArray array.
// Eg, assume nn = 5th trial, dataSizeArray[4] = 30000
// We pick a number from a random location in values[0->30000]
// and time how long to takes to retrieve it.
// For each interval we store all the trial times, then
// comput their average and standard deviation.
// Binary Search Tree
index = random.Next (dataSizeArray[nn]);
pt.Start();
bt.findSymbol (values[index]);
pt.Stop();
timeBt[k] = pt.DurationSeconds;
// Hash Table
index = random.Next (dataSizeArray[nn]);
pt.Start();
ht[values[index]].ToString();
pt.Stop();
timeHt[k] = pt.DurationSeconds;
}
// Compute the mean time
for (int i=0; i<trials; i++) {
meanBt = meanBt + timeBt[i];
meanHt = meanHt + timeHt[i];
}
meanBt = meanBt/trials;
meanHt = meanHt/trials;
Console.WriteLine();
Console.WriteLine ("\nAverage Time for Binary Tree = {0}", meanBt);
// Compute standard deviation
double sd = 0;
for (int i=0; i<trials; i++)
sd = sd + (timeBt[i] - meanBt)*(timeBt[i] - meanBt);
sd = Math.Sqrt (sd/trials);
// CV = coefficient of variation
Console.WriteLine ("Standard deviation = {0}, CV = {1}", sd, sd/meanBt);
Console.WriteLine ("\nAverage time for Hash Table = {0}", meanHt);
sd = 0;
for (int i=0; i<trials; i++)
sd = sd + (timeHt[i] - meanHt)*(timeHt[i] - meanHt);
sd = Math.Sqrt (sd/trials);
Console.WriteLine ("Standard deviation = {0}, CV = {1}", sd, sd/meanHt);
tw.WriteLine ("{0} {1} {2}", dataSizeArray[nn], meanBt, meanHt);
}
} finally {
tw.Close();
}
// Deletion tests
Console.WriteLine ("Test Deletion method\n");
bt = new TBinarySTree ();
bt.insert ("50", 50);
bt.insert ("60", 60);
bt.insert ("40", 40);
bt.insert ("30", 30);
bt.insert ("20", 20);
bt.insert ("35", 35);
bt.insert ("45", 45);
bt.insert ("44", 44);
bt.insert ("46", 46);
Console.WriteLine ("Number of nodes in the tree = {0}\n", bt.count());
Console.WriteLine ("Original: " + bt.drawTree());
bt.delete ("40");
Console.WriteLine ("Delete node 40: " + bt.drawTree());
bt.delete ("45");
Console.WriteLine ("Delete node 45: " + bt.drawTree());
Console.WriteLine ("\nSimple one layered tree");
bt = new TBinarySTree ();
bt.insert ("50", 50);
bt.insert ("20", 20);
bt.insert ("90", 90);
Console.WriteLine ("\nOriginal: " + bt.drawTree());
bt.delete ("50");
Console.WriteLine ("Delete node 50: " + bt.drawTree());
bt = new TBinarySTree ();
bt.insert ("50", 50);
bt.insert ("20", 20);
bt.insert ("90", 90);
Console.WriteLine ("\nOriginal: " + bt.drawTree());
bt.delete ("20");
Console.WriteLine ("Delete node 20: " + bt.drawTree());
bt = new TBinarySTree ();
bt.insert ("50", 50);
bt.insert ("20", 20);
bt.insert ("90", 90);
Console.WriteLine ("\nOriginal: " + bt.drawTree());
bt.delete ("90");
Console.WriteLine ("Delete node 90: " + bt.drawTree());
bt.delete ("20");
Console.WriteLine ("Delete node 20: " + bt.drawTree());
bt.delete ("50");
Console.WriteLine ("Delete node 50: " + bt.drawTree());
Console.WriteLine ("\n");
bt = new TBinarySTree ();
bt.insert ("L", 1);
bt.insert ("D", 2);
bt.insert ("C", 3);
bt.insert ("A", 4);
bt.insert ("H", 5);
bt.insert ("F", 6);
bt.insert ("J", 7);
bt.insert ("P", 8);
Console.WriteLine ("Original: " + bt.drawTree());
bt.delete ("J");
Console.WriteLine ("Delete J: " + bt.drawTree());
bt.delete ("C");
Console.WriteLine ("Delete C: " + bt.drawTree());
bt.delete ("L");
Console.WriteLine ("Delete L: " + bt.drawTree());
bt.delete ("D");
Console.WriteLine ("Delete D: " + bt.drawTree());
bt.delete ("A");
Console.WriteLine ("Delete A: " + bt.drawTree());
Console.ReadLine();
}