Comments on: Allpairs, Pairwise, Combinatorial Analysis http://www.testingmentor.com/imtesty/2009/11/11/allpairs-pairwise-combinatorial-analysis/ Treatises on the practice of software testing Tue, 07 Feb 2012 09:15:11 -0500 hourly 1 http://wordpress.org/?v=3.1.3 By: testingmentor http://www.testingmentor.com/imtesty/2009/11/11/allpairs-pairwise-combinatorial-analysis/comment-page-1/#comment-56 testingmentor Wed, 11 Nov 2009 18:29:46 +0000 http://testingmentor.com/imtesty/2009/11/11/allpairs-pairwise-combinatorial-analysis/#comment-56 Hi Krishtheindian, There are already numerous studies on the web and at http://www.pairwise.org. I don't understand what you mean by "compare both of them to see how the data is distributed." Combinatorial analysis is a technique for reducing the number of tests from all possible combinations when analyzing the interactions of varaible state combinations for n-way direct or semi-coupled unordered parameters. This has nothing to do with distribution of data. Exactly how are you going to produce manual tests to compare against the output of a tool? Producing all possible combinations of variable states is only reasonable for small feature sets because the total number of possible combinations is the Cartesian product of the number of variable states for all semi-coupled or dependent paramter. So, for trivial feature sets of say up to a hundred or so possible combinations comparisons between manaul tests and n-way analysis is relatively easy and has been done several times. But such trivial feature sets do not expose the full potentional of the toolset. Manual selection of tests from a feature set that produces say 10,000 or >1,000,000 possible combinations is unreasonably improbable. Friday, October 31, 2008 1:06 PM by I.M.Testy Hi Krishtheindian,

There are already numerous studies on the web and at http://www.pairwise.org.

I don’t understand what you mean by “compare both of them to see how the data is distributed.”

Combinatorial analysis is a technique for reducing the number of tests from all possible combinations when analyzing the interactions of varaible state combinations for n-way direct or semi-coupled unordered parameters.

This has nothing to do with distribution of data.

Exactly how are you going to produce manual tests to compare against the output of a tool?

Producing all possible combinations of variable states is only reasonable for small feature sets because the total number of possible combinations is the Cartesian product of the number of variable states for all semi-coupled or dependent paramter.

So, for trivial feature sets of say up to a hundred or so possible combinations comparisons between manaul tests and n-way analysis is relatively easy and has been done several times. But such trivial feature sets do not expose the full potentional of the toolset. Manual selection of tests from a feature set that produces say 10,000 or >1,000,000 possible combinations is unreasonably improbable.

Friday, October 31, 2008 1:06 PM by I.M.Testy

]]> By: testingmentor http://www.testingmentor.com/imtesty/2009/11/11/allpairs-pairwise-combinatorial-analysis/comment-page-1/#comment-55 testingmentor Wed, 11 Nov 2009 18:29:31 +0000 http://testingmentor.com/imtesty/2009/11/11/allpairs-pairwise-combinatorial-analysis/#comment-55 Hi, Recently I was searching for pairwise material and landed in your blog. Thanks a lot for pointing out PICT. it is Clear Introduction. Currently I am doing a study on the effectiveness of such a tool over manual test data generation. After producing the data manually and using the tool, I want to compare both of them to see how the data is distributed. Is there a methodical way to do this comparison? Any pointers will be greatly appriciated. Thanks in advance. Friday, October 31, 2008 11:02 AM by krishtheindian Hi,

Recently I was searching for pairwise material and landed in your blog. Thanks a lot for pointing out PICT. it is Clear Introduction.

Currently I am doing a study on the effectiveness of such a tool over manual test data generation. After producing the data manually and using the tool, I want to compare both of them to see how the data is distributed.

Is there a methodical way to do this comparison? Any pointers will be greatly appriciated. Thanks in advance.

Friday, October 31, 2008 11:02 AM by krishtheindian

]]> By: testingmentor http://www.testingmentor.com/imtesty/2009/11/11/allpairs-pairwise-combinatorial-analysis/comment-page-1/#comment-54 testingmentor Wed, 11 Nov 2009 18:29:20 +0000 http://testingmentor.com/imtesty/2009/11/11/allpairs-pairwise-combinatorial-analysis/#comment-54 A Good post BJ -- finally and after a long time- one that does not mention, compare or attack ET. It would be great if you blog on following topics in a similar fashion - I and many others are very eagerly watching your blog for the posts related to formal techniques. I especially watch to check how easy or difficult are these to learn and practice. 1. Boundary value analysis 2. EQ classes 3. Decision Tables 4. State machines 5. Risk based testing 6. Use case based techniques and so on... Shrini Tuesday, October 31, 2006 9:25 AM by Shrini A Good post BJ — finally and after a long time- one that does not mention, compare or attack ET.

It would be great if you blog on following topics in a similar fashion – I and many others are very eagerly watching your blog for the posts related to formal techniques. I especially watch to check how easy or difficult are these to learn and practice.

1. Boundary value analysis

2. EQ classes

3. Decision Tables

4. State machines

5. Risk based testing

6. Use case based techniques and so on…

Shrini

Tuesday, October 31, 2006 9:25 AM by Shrini

]]>