public void TesteID3()
{
Attribute ceu = new Attribute("ceu", new string[] { "sol", "nublado", "chuva" });
Attribute temperatura = new Attribute("temperatura", new string[] { "alta", "baixa", "suave" });
Attribute humidade = new Attribute("humidade", new string[] { "alta", "normal" });
Attribute vento = new Attribute("vento", new string[] { "sim", "nao" });
Attribute[] attributes = new Attribute[] { ceu, temperatura, humidade, vento };
DataTable samples = getDataTable();
DecisionTreeID3 id3 = new DecisionTreeID3();
TreeNode root = id3.mountTree(samples, "result", attributes);
printNode(root, "");
}
// Use this for initialization
IEnumerator LoadFromDb()
{
using (SqliteConnection conn = new SqliteConnection(GetConnectionString()))
{
conn.Open();
SqliteCommand command = new SqliteCommand(conn);
id3 = new DecisionTreeID3();
int iBatch = 0;
foreach (var info in decisionTrees)
{
string query = string.Format("SELECT * FROM {0}", info.tableName);
command.CommandText = query;
foreach (var split in info.outcomes)
{
var reader = command.ExecuteReader();
DataTable table = new DataTable(info.tableName);
for (int i = 0; i < reader.FieldCount; i++)
{
if (reader.GetName(i) == info.predictColumn)
{
table.Columns.Add(reader.GetName(i), typeof(bool));
}
else
{
table.Columns.Add(reader.GetName(i), reader.GetFieldType(i));
}
}
while (reader.Read())
{
var dr = table.NewRow();
for (int i = 0; i < reader.FieldCount; i++)
{
if (table.Columns[i].ColumnName == info.predictColumn)
{
if (!reader.IsDBNull(i))
{
switch (split.type)
{
case DecisionTreeOutcome.OutcomeType.Int:
dr[i] = (reader.GetInt32(i) == split.value);
break;
case DecisionTreeOutcome.OutcomeType.Text:
dr[i] = (reader.GetString(i) == split.textValue);
break;
}
}
}
else
{
dr[i] = reader[i];
}
}
table.Rows.Add(dr);
if (++iBatch % batchSize == 0)
{
yield return(new WaitForEndOfFrame());
}
}
table.AcceptChanges();
split.root = id3.MountTree(table, info.predictColumn, info.GetAttributes());
Hawk.Log(string.Format("DECISION TREE {0}{1}\n{2}", info.decisionTreeName, split.name, split.root.PrintNode("--->")));
foreach (var obj in eventListeners)
{
obj.SendMessage("OnDecisionTreeReady", split.root);
}
reader.Close();
}
}
}
}
static void Main(string[] args)
{
/***
Множество значений атрибутов про сердце
***/
/*
string fileName = "data_heart.txt";
int countAtr = 22;
Attribute f1 = new Attribute("f1", new string[] {"0", "1"});
Attribute f2 = new Attribute("f2", new string[] {"0", "1"});
Attribute f3 = new Attribute("f3", new string[] {"0", "1"});
Attribute f4 = new Attribute("f4", new string[] {"0", "1"});
Attribute f5 = new Attribute("f5", new string[] {"0", "1"});
Attribute f6 = new Attribute("f6", new string[] {"0", "1"});
Attribute f7 = new Attribute("f7", new string[] {"0", "1"});
Attribute f8 = new Attribute("f8", new string[] {"0", "1"});
Attribute f9 = new Attribute("f9", new string[] {"0", "1"});
Attribute f10 = new Attribute("f10", new string[] {"0", "1"});
Attribute f11= new Attribute("f11", new string[] {"0", "1"});
Attribute f12= new Attribute("f12", new string[] {"0", "1"});
Attribute f13= new Attribute("f13", new string[] {"0", "1"});
Attribute f14= new Attribute("f14", new string[] {"0", "1"});
Attribute f15= new Attribute("f15", new string[] {"0", "1"});
Attribute f16= new Attribute("f16", new string[] {"0", "1"});
Attribute f17= new Attribute("f17", new string[] {"0", "1"});
Attribute f18= new Attribute("f18", new string[] {"0", "1"});
Attribute f19= new Attribute("f19", new string[] {"0", "1"});
Attribute f20= new Attribute("f20", new string[] {"0", "1"});
Attribute f21= new Attribute("f21", new string[] {"0", "1"});
Attribute f22= new Attribute("f22", new string[] {"0", "1"});
Attribute[] attributes = new Attribute[] { f1,f2,f3,f4,f5,f6,f7,f8,f9,f10,f11,f12,f13,f14,f15,f16,f17,f18,f19,f20,f21,f22};
*/
/***
Множество значений атрибутов про шары
***/
string fileName = "data_balloons.txt";
int countAtr = 4;
Attribute color = new Attribute("color", new string[] {"YELLOW", "GREEN"});
Attribute size = new Attribute("size", new string[] {"LARGE", "SMALL"});
Attribute act = new Attribute("act", new string[] { "STRETCH", "DIP"});
Attribute age = new Attribute("age", new string[] { "ADULT", "CHILD" });
Attribute[] attributes = new Attribute[] { color, size, act, age};
DataTable samples = getDataTable(fileName, countAtr);
DecisionTreeID3 id3 = new DecisionTreeID3();
TreeNode root = id3.mountTree(samples, "result", attributes);
StreamWriter writer = new StreamWriter ("out.txt");
printNode(root, "", writer);
writer.Close();
Console.WriteLine("Дерево в файле out.txt. Нажмите Enter для завершения");
Console.ReadLine();
}
private TreeNode InternalMountTree(DataTable samples, String targetCustomAttribute, CustomAttribute[] attributes)
{
if (AllSamplesPositives(samples, targetCustomAttribute) == true)
{
return(new TreeNode(new CustomAttribute(true)));
}
if (AllSamplesNegatives(samples, targetCustomAttribute) == true)
{
return(new TreeNode(new CustomAttribute(false)));
}
if (attributes.Length == 0)
{
return(new TreeNode(new CustomAttribute(GetMostCommonValue(samples, targetCustomAttribute))));
}
mTotal = samples.Rows.Count;
mTargetCustomAttribute = targetCustomAttribute;
mTotalPositives = CountTotalPositives(samples);
mEntropySet = CalcEntropy(mTotalPositives, mTotal - mTotalPositives);
CustomAttribute bestCustomAttribute = GetBestCustomAttribute(samples, attributes);
if (bestCustomAttribute == null)
{
return(new TreeNode(new CustomAttribute(false)));
}
TreeNode root = new TreeNode(bestCustomAttribute);
DataTable aSample = samples.Clone();
foreach (String value in bestCustomAttribute.values)
{
aSample.Rows.Clear();
DataRow[] rows = samples.Select(bestCustomAttribute.CustomAttributeName + " = " + "'" + value + "'");
foreach (DataRow row in rows)
{
aSample.Rows.Add(row.ItemArray);
}
ArrayList aCustomAttributes = new ArrayList(attributes.Length - 1);
for (int i = 0; i < attributes.Length; i++)
{
if (attributes[i].CustomAttributeName != bestCustomAttribute.CustomAttributeName)
{
aCustomAttributes.Add(attributes[i]);
}
}
if (aSample.Rows.Count == 0)
{
return(new TreeNode(new CustomAttribute(GetMostCommonValue(aSample, targetCustomAttribute))));
}
else
{
DecisionTreeID3 dc3 = new DecisionTreeID3();
TreeNode ChildNode = dc3.MountTree(aSample, targetCustomAttribute, (CustomAttribute[])aCustomAttributes.ToArray(typeof(CustomAttribute)));
root.AddTreeNode(ChildNode, value);
}
}
return(root);
}
private TreeNode internalMountTree(DataTable samples, string targetAttribute, Attribute[] attributes)
{
if (allSamplesPositives(samples, targetAttribute) == true)
return new TreeNode(new Attribute(true));
if (allSamplesNegatives(samples, targetAttribute) == true)
return new TreeNode(new Attribute(false));
if (attributes.Length == 0)
return new TreeNode(new Attribute(getMostCommonValue(samples, targetAttribute)));
mTotal = samples.Rows.Count;
mTargetAttribute = targetAttribute;
mTotalPositives = countTotalPositives(samples);
mEntropySet = calcEntropy(mTotalPositives, mTotal - mTotalPositives);
Attribute bestAttribute = getBestAttribute(samples, attributes);
TreeNode root = new TreeNode(bestAttribute);
DataTable aSample = samples.Clone();
foreach (string value in bestAttribute.values)
{
aSample.Rows.Clear();
DataRow[] rows = samples.Select(bestAttribute.AttributeName + " = " + "'" + value + "'");
foreach (DataRow row in rows)
{
aSample.Rows.Add(row.ItemArray);
}
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]);
}
if (aSample.Rows.Count == 0)
{
//return new TreeNode(new Attribute(getMostCommonValue(aSample, targetAttribute)));
return new TreeNode(new Attribute(null));
}
else
{
DecisionTreeID3 dc3 = new DecisionTreeID3();
TreeNode ChildNode = dc3.mountTree(aSample, targetAttribute, (Attribute[])aAttributes.ToArray(typeof(Attribute)));
root.AddTreeNode(ChildNode, value);
}
}
return root;
}