The Data Mining Forum                             open-source data mining software open-source data mining software data science journal data mining conferences
This forum is about data mining, data science and big data: algorithms, source code, datasets, implementations, optimizations, etc. You are welcome to post call for papers, data mining job ads, link to source code of data mining algorithms or anything else related to data mining. The forum is hosted by P. Fournier-Viger. No registration is required to use this forum!.  
Lift Ratio
Posted by: Fendi
Date: July 04, 2012 01:55AM

I am a student from Indonesia
I want to ask, how to find a lift ratio of fp-growth algorithm contained in spmfGUIv079?

formula to calculate the lift ratio = "confidence value / expected value of confidence"
where to find the expected confidence = number of transactions containing consequent items / total number of transactions

I am very please help

thank you

regards.

Options: ReplyQuote
Re: Lift Ratio
Date: July 04, 2012 06:53AM

Hi Fendi,

I will explain to you how to add the lift measure to SPMF.

It is not very complciated. The lift is calculated as lift( X --> Y) = sup(X U Y) / (sup(x)*sup(y)). Because the confidence is confidence( X --> Y) = sup(X U Y) / sup(x), we can calculate the lift as follows: lift(x -->Y) = confidence(X -->Y) / sup(y)

So I will use this formula to calculate the lift.

What we first need to do is to modify the class RuleAgrawal in the package ca.pfv.spmf.associationrules.agrawal_FPGrowth_version by adding this method to calculate the lift:

public double getLift(){
		return (double)getConfidence() / itemset2.getAbsoluteSupport();
	}

Then, we will modify the code from the method printRules() in the class RulesAgrawal so that the lift is shown in the result. You just need to add one line:

public void printRules(int objectsCount){
		System.out.println(" ------- " + name + " -------"winking smiley;
		int i=0;
		for(RuleAgrawal rule : rules){
			System.out.print("  rule " + i + ":  " + rule.toString());
			System.out.print("support :  " + rule.getRelativeSupport(objectsCount) +
					" (" + rule.getAbsoluteSupport() + "/" + objectsCount + "winking smiley "winking smiley;
			System.out.print("confidence :  " + rule.getConfidence());
			System.out.print(" lift :  " + rule.getLift());
			System.out.println(""winking smiley;
			i++;
		}
		System.out.println(" --------------------------------"winking smiley;
	}

After that, we will modify the class AlgoFPGrowth in the package ca.pfv.spmf.frequentpatterns.fpgrowth. We will add the method to sort the items in an itemset:

public void sortItemset(Itemset itemset){
		Collections.sort(itemset.getItems(), new Comparator<Integer>() {
			public int compare(Integer o1,Integer o2) {
				return o1 - o2;
			}
		});
	}

And we will add this line in the method addAllCombinationsForPathAndPrefix of AlgoFPGrowth::

private void addAllCombinationsForPathAndPrefix(FPNode node, Itemset prefix) {
		// We add the node to the prefix
		Itemset itemset = prefix.cloneItemset();
		itemset.addItem(node.itemID);

		itemset.setTransactioncount(node.counter); 

		sortItemset(itemset);
		frequentItemsets.addItemset(itemset, itemset.size());
		
		// recursive call if there is a node link
		if(node.nodeLink != null){
			addAllCombinationsForPathAndPrefix(node.nodeLink, prefix);
			addAllCombinationsForPathAndPrefix(node.nodeLink, itemset);
		}
	}

And we will add this line in the method fpgrowthMoreThanOnePath of AlgoFPGrowth:

private void fpgrowthMoreThanOnePath(FPTree tree, Itemset prefixAlpha, Map<Integer, Integer> mapSupport) {
		// We process each frequent item in the header table list of the tree in reverse order.
		for(int i= tree.headerList.size()-1; i>=0; i--){
			Integer item = tree.headerList.get(i);
			
			int support = mapSupport.get(item);
			// if the item is not frequent, we skip it
			if(support <  relativeMinsupp){
				continue;
			}
			// Create Beta by concatening Alpha with the current item
			// and add it to the list of frequent patterns
			Itemset beta = prefixAlpha.cloneItemset();
			beta.addItem(item);
			if(prefixAlpha.getAbsoluteSupport() < support){
				beta.setTransactioncount(prefixAlpha.getAbsoluteSupport());
			}else{
				beta.setTransactioncount(support);
			}
			
			sortItemset(beta);
			frequentItemsets.addItemset(beta, beta.size());
			
			// === Construct beta's conditional pattern base ===
			// It is a subdatabase which consists of the set of prefix paths
			// in the FP-tree co-occuring with the suffix pattern.
			List<List<FPNode>> prefixPaths = new ArrayList<List<FPNode>>();
			FPNode path = tree.mapItemNodes.get(item);
			while(path != null){
				// if the path is not just the root node
				if(path.parent.itemID != -1){
					// create the prefixpath
					List<FPNode> prefixPath = new ArrayList<FPNode>();
					// add this node.
					prefixPath.add(path);   // NOTE: we add it just to keep its support,
					// actually it should not be part of the prefixPath
					
					//Recursively add all the parents of this node.
					FPNode parent = path.parent;
					while(parent.itemID != -1){
						prefixPath.add(parent);
						parent = parent.parent;
					}
					prefixPaths.add(prefixPath);
				}
				// We will look for the next prefixpath
				path = path.nodeLink;
			}
			
			// (A) Calculate the frequency of each item in the prefixpath
			Map<Integer, Integer> mapSupportBeta = new HashMap<Integer, Integer>();
			// for each prefixpath
			for(List<FPNode> prefixPath : prefixPaths){
				// the support of the prefixpath is the support of its first node.
				int pathCount = prefixPath.get(0).counter;  
				for(int j=1; j<prefixPath.size(); j++){  // for each node, except the first one, we count the frequency
					FPNode node = prefixPath.get(j);
					if(mapSupportBeta.get(node.itemID) == null){
						mapSupportBeta.put(node.itemID, pathCount);
					}else{
						mapSupportBeta.put(node.itemID, mapSupportBeta.get(node.itemID) + pathCount);
					}
				}
			}
			
			// (cool smiley Construct beta's conditional FP-Tree
			FPTree treeBeta = new FPTree();
			// add each prefixpath in the FP-tree
			for(List<FPNode> prefixPath : prefixPaths){
				treeBeta.addPrefixPath(prefixPath, mapSupportBeta, relativeMinsupp); 
			}  
			treeBeta.createHeaderList(mapSupportBeta); 
			
			// Mine recursively the Beta tree.
			if(treeBeta.root.childs.size() > 0){
				fpgrowth(treeBeta, beta, mapSupportBeta);
			}
		}
		
	}

and lastly, we need to add this line in the method code of AlgoAgrawalFaster94_FPGrowth_version:

private void apGenrules(int k, int m, Itemset lk, Set<Itemset> Hm) {
//		System.out.println(" " + lk.toString() + "  " + Hm.toString());
		if(k > m+1){
			Set<Itemset> Hm_plus_1 = generateCandidateSizeK(Hm);
			Set<Itemset> Hm_plus_1_for_recursion = new HashSet<Itemset>();
			for(Itemset hm_P_1 : Hm_plus_1){
				Itemset itemset_Lk_minus_hm_P_1 = lk.cloneItemSetMinusAnItemset(hm_P_1);

				calculateSupport(itemset_Lk_minus_hm_P_1);   // THIS COULD BE DONE ANOTHER WAY ?
				                                             // IT COULD PERHAPS BE IMPROVED....
				
				calculateSupport(hm_P_1);  // if we want to calculate the lift, we need to add this.
//				System.out.println(hm_P_1.getAbsoluteSupport());
				
				double conf = ((double)lk.getAbsoluteSupport()) / ((double)itemset_Lk_minus_hm_P_1.getAbsoluteSupport());
				
				if(conf >= minconf){
					RuleAgrawal rule = new RuleAgrawal(itemset_Lk_minus_hm_P_1, hm_P_1, lk.getAbsoluteSupport(), conf);
					rules.addRule(rule);
					Hm_plus_1_for_recursion.add(hm_P_1);
				}
			}
			apGenrules(k, m+1, lk, Hm_plus_1_for_recursion);
		}
	}

This latter line will calculate sup(Y) for each rule. For this to work, the itemset need to be sorted. This is why, i have added the sortItemset(..) method.

After that, if you run the test file MainTestAllAssociationRules_FPGrowth_version, the lift will be shown in the console:

 ------- All association rules -------
  rule 0:  5  ==> 4 support :  0.5 (3/6) confidence :  0.6 lift :  0.15
  rule 1:  4  ==> 5 support :  0.5 (3/6) confidence :  0.75 lift :  0.15
  rule 2:  1  ==> 4 support :  0.5 (3/6) confidence :  0.75 lift :  0.1875
  rule 3:  4  ==> 1 support :  0.5 (3/6) confidence :  0.75 lift :  0.1875
  rule 4:  2  ==> 4 support :  0.6666666666666666 (4/6) confidence :  0.6666666666666666 lift :  0.16666666666666666
  rule 5:  4  ==> 2 support :  0.6666666666666666 (4/6) confidence :  1.0 lift :  0.16666666666666666
  rule 6:  3  ==> 5 support :  0.5 (3/6) confidence :  0.75 lift :  0.15
  rule 7:  5  ==> 3 support :  0.5 (3/6) confidence :  0.6 lift :  0.15
  rule 8:  2  ==> 3 support :  0.6666666666666666 (4/6) confidence :  0.6666666666666666 lift :  0.16666666666666666
  rule 9:  3  ==> 2 support :  0.6666666666666666 (4/6) confidence :  1.0 lift :  0.16666666666666666
  rule 10:  1  ==> 5 support :  0.6666666666666666 (4/6) confidence :  1.0 lift :  0.2
  rule 11:  5  ==> 1 support :  0.6666666666666666 (4/6) confidence :  0.8 lift :  0.2
  rule 12:  1  ==> 2 support :  0.6666666666666666 (4/6) confidence :  1.0 lift :  0.16666666666666666
  rule 13:  2  ==> 1 support :  0.6666666666666666 (4/6) confidence :  0.6666666666666666 lift :  0.16666666666666666
  rule 14:  2  ==> 5 support :  0.8333333333333334 (5/6) confidence :  0.8333333333333334 lift :  0.16666666666666669
  rule 15:  5  ==> 2 support :  0.8333333333333334 (5/6) confidence :  1.0 lift :  0.16666666666666666
  rule 16:  2 5  ==> 4 support :  0.5 (3/6) confidence :  0.6 lift :  0.15
  rule 17:  2 4  ==> 5 support :  0.5 (3/6) confidence :  0.75 lift :  0.15
  rule 18:  4 5  ==> 2 support :  0.5 (3/6) confidence :  1.0 lift :  0.16666666666666666
  rule 19:  4  ==> 2 5 support :  0.5 (3/6) confidence :  0.75 lift :  0.15
  rule 20:  5  ==> 2 4 support :  0.5 (3/6) confidence :  0.6 lift :  0.15
  rule 21:  1 5  ==> 4 support :  0.5 (3/6) confidence :  0.75 lift :  0.1875
  rule 22:  1 4  ==> 5 support :  0.5 (3/6) confidence :  1.0 lift :  0.2
  rule 23:  4 5  ==> 1 support :  0.5 (3/6) confidence :  1.0 lift :  0.25
  rule 24:  5  ==> 1 4 support :  0.5 (3/6) confidence :  0.6 lift :  0.19999999999999998
  rule 25:  1  ==> 4 5 support :  0.5 (3/6) confidence :  0.75 lift :  0.25
  rule 26:  4  ==> 1 5 support :  0.5 (3/6) confidence :  0.75 lift :  0.1875
  rule 27:  1 2  ==> 4 support :  0.5 (3/6) confidence :  0.75 lift :  0.1875
  rule 28:  1 4  ==> 2 support :  0.5 (3/6) confidence :  1.0 lift :  0.16666666666666666
  rule 29:  2 4  ==> 1 support :  0.5 (3/6) confidence :  0.75 lift :  0.1875
  rule 30:  1  ==> 2 4 support :  0.5 (3/6) confidence :  0.75 lift :  0.1875
  rule 31:  4  ==> 1 2 support :  0.5 (3/6) confidence :  0.75 lift :  0.1875
  rule 32:  2 3  ==> 5 support :  0.5 (3/6) confidence :  0.75 lift :  0.15
  rule 33:  2 5  ==> 3 support :  0.5 (3/6) confidence :  0.6 lift :  0.15
  rule 34:  3 5  ==> 2 support :  0.5 (3/6) confidence :  1.0 lift :  0.16666666666666666
  rule 35:  5  ==> 2 3 support :  0.5 (3/6) confidence :  0.6 lift :  0.15
  rule 36:  3  ==> 2 5 support :  0.5 (3/6) confidence :  0.75 lift :  0.15
  rule 37:  1 2  ==> 5 support :  0.6666666666666666 (4/6) confidence :  1.0 lift :  0.2
  rule 38:  1 5  ==> 2 support :  0.6666666666666666 (4/6) confidence :  1.0 lift :  0.16666666666666666
  rule 39:  2 5  ==> 1 support :  0.6666666666666666 (4/6) confidence :  0.8 lift :  0.2
  rule 40:  5  ==> 1 2 support :  0.6666666666666666 (4/6) confidence :  0.8 lift :  0.2
  rule 41:  2  ==> 1 5 support :  0.6666666666666666 (4/6) confidence :  0.6666666666666666 lift :  0.16666666666666666
  rule 42:  1  ==> 2 5 support :  0.6666666666666666 (4/6) confidence :  1.0 lift :  0.2
  rule 43:  1 2 5  ==> 4 support :  0.5 (3/6) confidence :  0.75 lift :  0.1875
  rule 44:  1 2 4  ==> 5 support :  0.5 (3/6) confidence :  1.0 lift :  0.2
  rule 45:  1 4 5  ==> 2 support :  0.5 (3/6) confidence :  1.0 lift :  0.16666666666666666
  rule 46:  2 4 5  ==> 1 support :  0.5 (3/6) confidence :  1.0 lift :  0.25
  rule 47:  2 4  ==> 1 5 support :  0.5 (3/6) confidence :  0.75 lift :  0.1875
  rule 48:  1 2  ==> 4 5 support :  0.5 (3/6) confidence :  0.75 lift :  0.25
  rule 49:  1 5  ==> 2 4 support :  0.5 (3/6) confidence :  0.75 lift :  0.1875
  rule 50:  1 4  ==> 2 5 support :  0.5 (3/6) confidence :  1.0 lift :  0.2
  rule 51:  4 5  ==> 1 2 support :  0.5 (3/6) confidence :  1.0 lift :  0.25
  rule 52:  2 5  ==> 1 4 support :  0.5 (3/6) confidence :  0.6 lift :  0.19999999999999998
  rule 53:  4  ==> 1 2 5 support :  0.5 (3/6) confidence :  0.75 lift :  0.1875
  rule 54:  1  ==> 2 4 5 support :  0.5 (3/6) confidence :  0.75 lift :  0.25
  rule 55:  5  ==> 1 2 4 support :  0.5 (3/6) confidence :  0.6 lift :  0.19999999999999998
 --------------------------------

For example, if you consider the rule 4 ==> 1 2 5, the lift is calculated as lift(4--> 1 2 5) = sup(4 1 2 5) / (sup(4)* sup( 1 2 5) = 3 / (4*4) = 0.1875.

If you want to add the lift to the version of FPGrowth that save the result to a file, the idea should be the same, except that you will have to modify files in the package "ca.pfv.spmf.associationrules.agrawal_FPGrowth_version_saveToFile".

Hope this helps!

Philippe



Edited 7 time(s). Last edit at 07/04/2012 07:53AM by webmasterphilfv.

Options: ReplyQuote
Re: Lift Ratio
Date: July 04, 2012 08:56AM

I have posted a new version of SPMF with the lift on the website (0.83). You can download it and try it out!

Best,

Philippe



Edited 1 time(s). Last edit at 07/04/2012 09:01AM by webmasterphilfv.

Options: ReplyQuote
Re: Lift Ratio
Posted by: Fendi
Date: July 05, 2012 11:36PM

thank you very much for your response
Your explanation really helped me
I pray that God always bless you
once again thank you very much

can I ask again .. how to translate the results of the association rule from fp-algorithm using graphics ..?
make it easier for the user in understanding the pattern

thank you..
regards.

Options: ReplyQuote
Re: Lift Ratio
Date: July 06, 2012 04:14PM

Hi Fendi,

You are welcome. I'm also happy about your suggestion about the lift because it help me to improve the software.

For visualizing the association rules, I don't have anything for that. Currently, the ouptut is just the console or a file.

There are some researchers that have worked on how to display the result of association rule mining. For example, here there is some Java code:

http://www2.lifl.fr/~jourdan/download/arv.html

But I have not tried it and they use a different format as me. But maybe that you could convert to their format or modify my code to write to their format. It would be interesting to try this out. EDIT: I have checked a little bit more, and I think that they don't use the standard definition of an association rule so it may not work.

There are also some other works on association rule visualization that you can find by searching "association rule visualization" in Google. Maybe you can get some code that you could reuse.

Actually, the most appropriate way to visualize the rule may depend on your application. Maybe you could design something specific that make more sense for your application.

I just give you some ideas.

Best,

Philippe



Edited 1 time(s). Last edit at 07/06/2012 04:20PM by webmasterphilfv.

Options: ReplyQuote
Re: Lift Ratio
Posted by: Fendi
Date: July 11, 2012 01:36AM

once again thank you very much for your guidance ..
I hope to continue to learn from you ..
I'm very interested in the Data Minning particular association rule

I hope you do not get bored with my questions ..
I want to learn more about fp-growth algorithm

Can you please tell me how to step in building an application using fp-growth algorithm to find frequent itemsets and association rules on the application that you build from the java programming language

thanks again for your help

regards

Options: ReplyQuote
Re: Lift Ratio
Date: July 11, 2012 08:02PM

Hi Fendi,

You are welcome.

To build an application using my implementation of FPGrowth, it is quite simple.

First you need to download the source code version of SPMF from the download page (I think that you have done this already) and import my code in your project.

Then, you should check the examples about how to use the algorithms. In the documentation on the website, there is the list of examples about how to use each algorithm. There is also a description of each example in the documentation.

So for example, if you are interested by generating association rules with the lift, you should check example #20. This example tells you that this example correspond the file "MainTestAllAssociationRules_FPGrowth_version_with_lift.java" in the source code. If you open this file, you will the code for executing this example.

So you should check the code for these examples. The code is generally very simple. Then you should copy the code from the example in your application and modify it according to your needs. For example, you could change the parameters and the input file.

Hope this helps. Best,

Philippe



Edited 1 time(s). Last edit at 07/11/2012 08:05PM by webmasterphilfv.

Options: ReplyQuote
Re: Lift Ratio
Posted by: Fendi
Date: July 12, 2012 01:44AM

thank you very much for your guidance ..thumbs up

I hope to continue to learn from you ..

want to ask again ... smiling smiley

whether it could result from printstat fp-growth association rule is displayed in the notepad .. ??

Options: ReplyQuote
Re: Lift Ratio
Date: July 12, 2012 10:59AM

Yes. it can. actually, what you would need to do is to take the version of FPGrowth that save the result to file and make sure that it saves the result to a ".txt" file.

Then you would use just add some java code to launch the default text editor (such as notepad) to open the file.

To open a file with notepad, you could use something like that:

String outputFile ="C:\\ .... the name of your file";

if (desktop.isSupported(Desktop.Action.OPEN)) {
					  try{
					  desktop.open(new File(outputFile));
					  }catch(Exception e){
						  JOptionPane.showMessageDialog(null,
								    "An error occured while opening the output file. ERROR MESSAGE = " + e.toString(), "Error",
								    JOptionPane.ERROR_MESSAGE);
					  }
				  }

It will open notepad if the default text editor is notepad. That's it.


Note that if you use the version of FPGrowth that keep the result into memory instead (for example to sort the rules by confidence), then you would need to modify it so that it save to a file, which is not complicated.

This is just the basic idea.

Philippe



Edited 1 time(s). Last edit at 07/12/2012 12:47PM by webmasterphilfv.

Options: ReplyQuote
Re: Lift Ratio
Posted by: Fendi
Date: July 12, 2012 08:53PM

I'm referring to is the method of

public void printStats () {


System.out.println ("FP-GROWTH ============= - STATS ============="winking smiley;
long temps = endtime - startTimestamp;
System.out.println ("Transactions count from database:" + contextSize);
System.out.println ("Frequent itemsets count:" + frequentItemsets.getItemsetsCount ());
System.out.println ("Total time ~" + temps + "ms"winking smiley;
System.out.println ("=========================================== ======== "winking smiley;

and

public void printStats () {
System.out
. System.out.println ("============= ASSOCIATION RULE GENERATION - STATS ============="winking smiley;
System.out.println ("Number of association rules generated:" + ruleCount);
System.out.println ("Total time ~" + (endTimeStamp - startTimestamp) + "ms"winking smiley;
System.out
. System.out.println ("============================================== ===== "winking smiley;

can be displayed together with the association rule algorithm which uses fp-growth on the notepad .. ??

Options: ReplyQuote
Re: Lift Ratio
Posted by: Fendi
Date: July 12, 2012 09:21PM

I mean above, is ..
What could be the result of the generated rule in fp-growth algorithm combined with the method of

public void printStats () {


System.out.println ("FP-GROWTH ============= - ============= STATS" Winking smiley;
long temps = endtime - startTimestamp;
System.out.println ("Transactions count from database:" + contextSize);
System.out.println ("Frequent itemsets count:" + frequentItemsets.getItemsetsCount ());
System.out.println ("Total time ~" + temps + "ms" Winking smiley;
System.out.println ("=========================================== ======== "Winking smiley;

and

public void printStats () {
System.out
. System.out.println ("============= ASSOCIATION RULE GENERATION - STATS =============" Winking smiley;
System.out.println ("Number of association rules generated:" + ruleCount);
System.out.println ("Total time ~" + (endTimeStamp - startTimestamp) + "ms" Winking smiley;
System.out
. System.out.println ("=========================================== ======== "Winking smiley;

thank you very much..

regards.

Options: ReplyQuote
Re: Lift Ratio
Date: July 13, 2012 05:15AM

Ok I understand your question now.

First you could add this code in the class "AlgoAgrawalFaster94_FPGrowth_version_saveToFile" :

public void saveTextToFile(String text) throws IOException{
		writer.write(text);
	}
	
	public void saveStatsToFile() throws IOException {
		saveTextToFile("=============  ASSOCIATION RULE GENERATION - STATS =============\n"winking smiley;
		saveTextToFile(" Number of itemsets generated : "
				+ patterns.getItemsetsCount() + "\n"winking smiley;
		saveTextToFile(" Number of association rules generated : "
				+ ruleCount + "\n"winking smiley;
		saveTextToFile(" Total time ~ " + (endTimeStamp - startTimestamp)
				+ " ms\n"winking smiley;
		saveTextToFile("===================================================\n"winking smiley;
	}

Then, you would replace this code in the same file:

		// close the file
		writer.close();
		endTimeStamp = System.currentTimeMillis();

with:

                saveStatsToFile();
		// close the file
		writer.close();
		endTimeStamp = System.currentTimeMillis();

Then it should work.

Note that here I just print the number of itemsets and the number of association rules and the time for association rule generation. I did not print the time for itemset generation. But it could be done by passing the information from the class that perform itemset generation to the class that finds the rules.

Best,

Philippe



Edited 1 time(s). Last edit at 07/13/2012 05:17AM by webmasterphilfv.

Options: ReplyQuote
Re: Lift Ratio
Posted by: Fendi
Date: July 14, 2012 12:23AM

many thank for you..
wish all the best for you smiling smiley

Options: ReplyQuote
Re: Lift Ratio
Posted by: Fendi
Date: July 16, 2012 03:04AM

excuse me .. I want to ask you again on
what should I do if I want to make combobox in netbeans JFrame Form in which there are only two options, namely fp-growth frequent itemsets and association rules fp-growth ..?
please help ..

regards

Options: ReplyQuote
Re: Lift Ratio
Date: July 16, 2012 10:46AM

Hi Fendi,

I don't know how to make it with the visual editor of NetBeans because I use Eclipse instead.

But you get go there : http://docs.oracle.com/javase/tutorial/uiswing/components/combobox.html to get some source code about how to use the combo box if you write the code manually.

For example, you can place the JComboBox on the JFrame and then write some code to set the elements in the comboBox. For example, something like that (taken from the previous website):

String[] petStrings = { "Bird", "Cat", "Dog", "Rabbit", "Pig" };

//Create the combo box, select item at index 4.
//Indices start at 0, so 4 specifies the pig.
JComboBox petList = new JComboBox(petStrings);
....
petList.setSelectedIndex(4);
petList.addActionListener(this);


You can check the Java code on the webpage that I suggested to see some demo about how to use the combo box and some source code.

Hope this helps,

Philippe

Options: ReplyQuote
Re: Lift Ratio
Posted by: Fendi
Date: July 19, 2012 02:33AM

Thank you verymuch for your help..

I want to ask again, is the result of the support and confidence in association rule algorithms fpgrowth, the decimal value is rounded to 2 decimal places

thank you.

Fendi

Options: ReplyQuote
Re: Lift Ratio
Posted by: Fendi
Date: July 19, 2012 03:54AM

I want to ask again, is the result of the support and confidence in association rule algorithms fpgrowth, the decimal value is rounded to 2 decimal places

example 3.138793526238352 to be 3.14

thank you.

Fendi

Options: ReplyQuote


This forum is powered by Phorum and provided by P. Fournier-Viger (© 2012).
Terms of use.