static void Main(string[] args)
{
//Outlook
Attribute outlook = new Attribute("Outlook",
new string[] { "sunny", "overcast", "rainy" });
//Temperature
Attribute temp = new Attribute("Temperature",
new string[] { "hot", "mild", "cool" });
//Humidity
Attribute humidity = new Attribute("Humidity",
new string[] { "high", "normal" });
//Windy
Attribute windy = new Attribute("Windy",
new string[] { "true", "false" });
//načíst data z CSV
Attribute[] attributes = new Attribute[] { outlook, temp, humidity, windy };
DataTable samples = loadCSV();
//ID3 na data
ID3 id3 = new ID3();
Node root = id3.makeTree(samples, "play", attributes);
Console.ReadKey();
printNode(root, "--");
Console.ReadKey();
}
private Node ID3Tree(DataTable DataSet, string Target, Attribute[] attributes)
{
//Entropie tabulky je 0.00 = JE LEAF
if (IsPositive(DataSet, Target) == true)
{
Node tmpNode = new Node(new Attribute(true));
tmpNode.totalData = DataSet.Rows.Count;
return(tmpNode);
}
if (IsNegative(DataSet, Target) == true)
{
Node tmpNode = new Node(new Attribute(false));
tmpNode.totalData = DataSet.Rows.Count;
return(tmpNode);
}
//OR
TotalLines = 14;
SearchFor = Target;
TrueNumber = Count(DataSet);
//vypočteme entropii setu
Console.WriteLine("++++++++++++SET++++++++++++++");
EntropyOfSet = Entropy(TrueNumber, TotalLines - TrueNumber);
Console.WriteLine("+++++++++++++++++++++++++++++");
//Najit "nejlepší atribut" - který má největší information gain v tabulce
Attribute bestAttribute = GetBestGain(DataSet, attributes);
Console.WriteLine("BEST GAIN IN SET: " + bestAttribute.AttributeName);
Console.WriteLine();
//Udělat z nejlepšího atributu kořen (node)
Node root = new Node(bestAttribute);
root.totalData = TotalLines;
DataTable aSample = DataSet.Clone();
//Cyklus pro každou možnou hodnotu kterou muže nabývat "Best" atribut
foreach (string value in bestAttribute.values)
{
//Všechny řádky atributu nabývající danou hodnotu
aSample.Rows.Clear();
DataRow[] rows = DataSet.Select(bestAttribute.AttributeName + " = " + "'" + value + "'");
foreach (DataRow row in rows)
{
aSample.Rows.Add(row.ItemArray);
}
// nový list atributu bez "best gain" attributu
ArrayList aAttributes = new ArrayList(attributes.Length - 1);
for (int i = 0; i < attributes.Length; i++)
{
if (attributes[i].AttributeName != bestAttribute.AttributeName)
{
aAttributes.Add(attributes[i]);
}
}
//Rekurze ID3 algoritmu na novou pod-tabulku (omezený DataSet)
//Vkládat node do nodu vytvořeného z best atributu
ID3 nextTree = new ID3();
Node ChildNode = nextTree.makeTree(aSample, Target, (Attribute[])aAttributes.ToArray(typeof(Attribute)));
root.AddTreeNode(ChildNode, value);
}
return(root);
}