Prashant-Jonny / accord

Automatically exported from code.google.com/p/accord
0 stars 0 forks source link

Levene Test - Producing incorrect P-value, but correct statistic #29

Closed GoogleCodeExporter closed 8 years ago

GoogleCodeExporter commented 8 years ago
What steps will reproduce the problem?
1. Used the following data:

       List<double[]> _samples = new List<double[]>();

        double[] A = { 250, 260, 230, 270 };
        double[] B = { 310, 330, 280, 360 };
        double[] C = { 250, 230, 220, 260 };
        double[] D = { 340, 270, 300, 320 };
        double[] E = { 250, 240, 270, 290 };
            _samples.Add(A);
            _samples.Add(B);
            _samples.Add(C);
            _samples.Add(D);
            _samples.Add(E);

     LeveneTest result = new LeveneTest(_samples.ToArray(), true);

What is the expected output? What do you see instead?

Expect Output for result: 
statistic= ~0.7247
pValue = ~0.5886
This was outputted by R (Library: lawstat, levene.test) and propriety 
statistics software.

Actual Output for result:
statistic= ~0.7247
pValue = ~0.8228

What version of the product are you using? On what operating system?

2.8.1

Please provide any additional information below

Data Works fine for the newly added Bartlett's test

Original issue reported on code.google.com by andrewhd...@gmail.com on 3 Jan 2013 at 5:14

GoogleCodeExporter commented 8 years ago
Thanks. The Levene test should be have been computed as a one tail, upper tail 
test. Will be fixed on next release.

Original comment by cesarso...@gmail.com on 4 Jan 2013 at 3:07

GoogleCodeExporter commented 8 years ago
This issue was closed by revision r439.

Original comment by cesarso...@gmail.com on 23 Feb 2013 at 3:31