commons-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Anirban Basu (JIRA)" <j...@apache.org>
Subject [jira] Created: (MATH-228) Feature request for one and two sample Kolmogorov-Smirnov test as well as Lilliefors test
Date Sun, 28 Sep 2008 11:43:44 GMT
Feature request for one and two sample Kolmogorov-Smirnov test as well as Lilliefors test
-----------------------------------------------------------------------------------------

                 Key: MATH-228
                 URL: https://issues.apache.org/jira/browse/MATH-228
             Project: Commons Math
          Issue Type: New Feature
    Affects Versions: 2.0
         Environment: All
            Reporter: Anirban Basu
            Priority: Minor
             Fix For: 2.0


It would be very helpful to have implementations of one and two sample [Kolmogorov-Smirnov
test|http://en.wikipedia.org/wiki/Kolmogorov-Smirnov_test] as well as [Lilliefors test|http://en.wikipedia.org/wiki/Lilliefors_test]
with MATLAB-style results in future versions of Commons Math.

For example, Lilliefors test on sample data:

sampleVector = [0.0033413022337048857, 0.008527692135731013, -0.004902763950955454, 0.033018433100296396,
-0.020495504044139023, 0.003978726052913162, 0.003847972673931109, 0.009160477945515444, -0.011113437653216639,
-0.01164235145079795, 0.017180306607011864, -0.01818483009998717, -0.010479811709006803, -0.033991339307749,
-0.007057160031600951, -1.2398497120424956E-4, 0.0026913151777877564, 0.03580425341677764,
-0.006404370278251359, 0.007579083257585828, -0.005912037207256193, 0.01241830354576745, -0.0012524631744377235,
-0.005900927958040758, 0.0028847985848513558, 0.005313417226899042, 0.018923743379700153,
0.010976836172447269, -0.017847220928846164, 0.0024067380689056783, -0.011912393656503872,
-0.019985462687391875, 0.017318878212931876, 0.003592873590795409, -0.00332615776078915, -0.018222673013956525,
-0.021591768336351125];

[h, p] = lillietest(sampleVector)
Warning: P is greater than the largest tabulated value, returning 0.5.
 			> In lillietest at 166
			h =
					0
			p =
					0.5000

This uses Lilliefors test for normality. The test returns that h=0, i.e. the null hypothesis
that the data vector obeys Normal distribution.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.


Mime
View raw message