public void TestMaxHeapFPGrowth()
{
FPGrowth<string> fp = new FPGrowth<string>();
List<KeyValuePair<List<string>, long>> transactions = new List<KeyValuePair<List<string>, long>>();
transactions.Add(new KeyValuePair<List<string>, long>(Arrays.asList("E", "A", "D", "B"), 1L));
transactions.Add(new KeyValuePair<List<string>, long>(Arrays.asList("D", "A", "C", "E", "B"), 1L));
transactions.Add(new KeyValuePair<List<string>, long>(Arrays.asList("C", "A", "B", "E"), 1L));
transactions.Add(new KeyValuePair<List<string>, long>(Arrays.asList("B", "A", "D"), 1L));
transactions.Add(new KeyValuePair<List<string>, long>(Arrays.asList("D"), 1L));
transactions.Add(new KeyValuePair<List<string>, long>(Arrays.asList("D", "B"), 1L));
transactions.Add(new KeyValuePair<List<string>, long>(Arrays.asList("A", "D", "E"), 1L));
transactions.Add(new KeyValuePair<List<string>, long>(Arrays.asList("B", "C"), 1L));
using (var outWriter = new StringWriter())
{
fp.GenerateTopKFrequentPatterns(
transactions,
fp.GenerateFList(transactions, 3),
3,
100,
new HashSet<string>(),
new StringOutputConverter(new SequenceFileOutputCollector<string, TopKStringPatterns>(outWriter, "{0}\t{1},")),
new ContextStatusUpdater());
var res = @"C ([B,C],3),"
+ @"E ([A,E],4), ([A,B,E],3), ([A,D,E],3),"
+ @"A ([A],5), ([A,D],4), ([A,E],4), ([A,B],4), ([A,B,E],3), ([A,D,E],3), ([A,B,D],3),"
+ @"D ([D],6), ([B,D],4), ([A,D],4), ([A,D,E],3), ([A,B,D],3),"
+ @"B ([B],6), ([A,B],4), ([B,D],4), ([A,B,D],3), ([A,B,E],3), ([B,C],3),";
Assert.AreEqual(outWriter.ToString(), res);
}
}
static void Main(string[] args)
{
/*
This sample demonstrate how-to use the FP Growth algorithm.
As a test data I use a data from an imagine book store:
1) We have a list of authors, which the store sells.
2) We have an information about shopping cart of some users.
As a result we get the following information:
1. Order by popularity of the authors/set of authors.
2. What we can advise the user, basing on his shopping cart. It is ordered by probability (frequent).
*/
TypeRegister.Register<BookAuthor>((a, b) => a.Name.CompareTo(b.Name));
var itemsets = new List<Itemset<BookAuthor>>();
var transactions = new SampleHelper().Transactions.ToList();
var fpGrowth = new FPGrowth<BookAuthor>
{
MinSupport = (double) 1/9,
SaveItemset = itemset => itemsets.Add(itemset),
GetTransactions = () => transactions
};
fpGrowth.ProcessTransactions();
var rules = new List<Rule<BookAuthor>>();
var agrawal = new AgrawalFaster<BookAuthor>
{
MinLift = 0.01,
MinConfidence = 0.01,
TransactionsCount = transactions.Count(),
GetItemsets = () => itemsets,
SaveRule = rule => rules.Add(rule)
};
agrawal.Run();
foreach (var item in itemsets.OrderByDescending(v => v.Support))
{
Console.WriteLine(string.Join("; ", item.Value.OrderBy(i => i.Name)) + " #SUP: " + item.Support);
}
Console.WriteLine("====");
foreach (var item in rules.OrderByDescending(r => r.Confidence))
{
Console.WriteLine(string.Join("; ", item.Combination) + " => " + string.Join("; ", item.Remaining) + " ===> Confidence: " + item.Confidence + " Lift: " + item.Lift);
}
Console.ReadLine();
}
public override void ViewDidLoad()
{
TypeRegister.Register<BookAuthor>((a, b) => a.Name.CompareTo(b.Name));
var itemsets = new List<Itemset<BookAuthor>>();
var transactions = new SampleHelper().Transactions.ToList();
var fpGrowth = new FPGrowth<BookAuthor>
{
MinSupport = (double) 1/9,
SaveItemset = itemset => itemsets.Add(itemset),
GetTransactions = () => transactions
};
fpGrowth.ProcessTransactions();
var rules = new List<Rule<BookAuthor>>();
var agrawal = new AgrawalFaster<BookAuthor>
{
MinLift = 0.01,
MinConfidence = 0.01,
TransactionsCount = transactions.Count(),
GetItemsets = () => itemsets,
SaveRule = rule => rules.Add(rule)
};
agrawal.Run();
FrequentItemsTable.RegisterClassForCellReuse (typeof(FrequentItemTableViewCell), FrequentTableViewDelegate.CellIdentifier);
FrequentItemsTable.Source = new FrequentTableViewDelegate{
GetData = () => itemsets.OrderByDescending(i=>i.Support).ToList()
};
FrequentRulesTable.RegisterClassForCellReuse (typeof(FrequentRulesTableViewCell), FrequentRulesTableViewDelegate.CellIdentifier);
FrequentRulesTable.Source = new FrequentRulesTableViewDelegate{
GetData = () => rules.OrderByDescending(i=>i.Confidence).ToList()
};
}
static void Main(string[] args)
{
DateTime start = DateTime.Now;
double minSupport = 0.005;
double minConfidence = 0.5;
string filepath = @"path/to/file";
List <Transaction> dataset = PrepareData(filepath);
FPGrowth method = new FPGrowth(dataset, minSupport * dataset.Count);
FPTree tree = method.GenerateTree(false);
var patterns = method.MineFrequentPatterns();
var rules = method.GenerateRules(patterns, minSupport, minConfidence);
DateTime end = DateTime.Now;
Console.WriteLine("Program finished in {0} seconds", (end - start).TotalSeconds);
patterns.ForEach(p => Console.WriteLine(p));
Console.WriteLine("==========================================================================");
rules.Sort((r1, r2) => r2.Confidence.CompareTo(r1.Confidence));
rules.ForEach(p => Console.WriteLine(p));
Console.ReadKey();
}
}//DMForm
//Methods *************************************************************
//Events **************************************************************
private void btnExecute_Click(object sender, EventArgs e)
{
if (cmbAlgorithm.SelectedItem == null)
{
MessageBox.Show("Please select an algorithm.");
return;
}//if no algorithm
string algorithm = cmbAlgorithm.SelectedItem.ToString();
if (algorithm.Equals("Apriori"))
{
rtbResults.Text = "";
List <ItemSetDto> results = new List <ItemSetDto>();
DateTime timer = DateTime.Now;
if (cmbSource.SelectedItem == null)
{
MessageBox.Show("Please select a data source.");
return;
}//if
if (txtMinSupport.Text.Equals(""))
{
MessageBox.Show("Please enter a Minimum Support");
return;
}//if
double minSupport = Convert.ToDouble(txtMinSupport.Text);
string datasource = cmbSource.SelectedItem.ToString();
DataSourceBLL dsBLL = new DataSourceBLL();
Apriori ap = new Apriori();
List <TransactionDto> transactions = dsBLL.process(datasource);
if (transactions == null)
{
return;
}
results = ap.Process(transactions, null, minSupport);
StringBuilder sb = new StringBuilder();
foreach (ItemSetDto isd in results)
{
sb.AppendLine(isd.toString());
}
rtbResults.AppendText(sb.ToString() + "\nTime to Complete: " + (DateTime.Now - timer).ToString());
}//if "Apriori"
else if (algorithm.Equals("FPGrowth"))
{
if (txtMinSupport.Text.Equals(""))
{
MessageBox.Show("Please enter a Minimum Support");
return;
}//if
double minSupport = Convert.ToDouble(txtMinSupport.Text);
string datasource = cmbSource.SelectedItem.ToString();
DataSourceBLL dsBLL = new DataSourceBLL();
List <TransactionDto> dtos = dsBLL.process(datasource);
if (dtos == null)
{
return;
}
rtbResults.Text = "";
DateTime timer = DateTime.Now;
FPGrowth fpg = new FPGrowth(dtos, minSupport);
List <FrequentPattern> fp = fpg.mine();
StringBuilder sb = new StringBuilder();
foreach (FrequentPattern f in fp)
{
sb.AppendLine(f.ToString());
}
rtbResults.AppendText(sb.ToString() + "\nTime to Complete: " + (DateTime.Now - timer).ToString());
}//if "FPGrowth"
else if (algorithm.Equals("Eclat"))
{
rtbResults.Text = "";
List <VerticalItemSetDto> results = new List <VerticalItemSetDto>();
DateTime timer = DateTime.Now;
if (cmbSource.SelectedItem == null)
{
MessageBox.Show("Please select a data source.");
return;
}//if
if (txtMinSupport.Text.Equals(""))
{
MessageBox.Show("Please enter a Minimum Support");
return;
}//if
double minSupport = Convert.ToDouble(txtMinSupport.Text);
string datasource = cmbSource.SelectedItem.ToString();
DataSourceBLL dsBLL = new DataSourceBLL();
Eclat ec = new Eclat();
List <TransactionDto> transactions = dsBLL.process(datasource);
if (transactions == null)
{
return;
}
results = ec.Process(transactions, null, minSupport);
StringBuilder sb = new StringBuilder();
foreach (VerticalItemSetDto isd in results)
{
sb.AppendLine(isd.toString());
}
rtbResults.AppendText(sb.ToString() + "\nTime to Complete: " + (DateTime.Now - timer).ToString());
} //if "Eclat"
} //btnExecute_Click
public void TestMaxHeapFPGrowthData2() {
FPGrowth<String> fp = new FPGrowth<String>();
ICollection<KeyValuePair<List<string>, long>> transactions = new List<KeyValuePair<List<string>, long>>();
transactions.Add(new KeyValuePair<List<string>, long>(Arrays.asList("X"), 12L));
transactions.Add(new KeyValuePair<List<string>, long>(Arrays.asList("Y"), 4L));
transactions.Add(new KeyValuePair<List<string>, long>(Arrays.asList("X", "Y"), 10L));
transactions.Add(new KeyValuePair<List<string>, long>(Arrays.asList("X", "Y", "Z"), 11L));
using (var outWriter = new StringWriter())
{
Console.WriteLine(fp.GenerateFList(transactions, 2));
fp.GenerateTopKFrequentPatterns(
transactions,
fp.GenerateFList(transactions, 2),
2,
100,
new HashSet<string>(),
new StringOutputConverter(new SequenceFileOutputCollector<string, TopKStringPatterns>(outWriter, "({0},{1}), ")),
new ContextStatusUpdater());
var res = @"(Z,([X,Y,Z],11)), (Y,([Y],25), ([X,Y],21), ([X,Y,Z],11)), (X,([X],33), ([X,Y],21), ([X,Y,Z],11)), ";
Assert.AreEqual(outWriter.ToString(), res);
}
}
