Represents a test that a sample comes from a specified distribution.
SystemObject Extreme.Statistics.TestsHypothesisTest Extreme.Statistics.TestsOneSampleTestDouble Extreme.Statistics.TestsOneSampleTest Extreme.Statistics.TestsOneSampleKolmogorovSmirnovTest
Namespace:
Extreme.Statistics.Tests
Assembly:
Extreme.Numerics (in Extreme.Numerics.dll) Version: 8.1.1
public sealed class OneSampleKolmogorovSmirnovTest : OneSampleTest
Public NotInheritable Class OneSampleKolmogorovSmirnovTest
Inherits OneSampleTest
public ref class OneSampleKolmogorovSmirnovTest sealed : public OneSampleTest
[<SealedAttribute>]
type OneSampleKolmogorovSmirnovTest =
class
inherit OneSampleTest
end
The OneSampleKolmogorovSmirnovTest type exposes the following members.
Top
Top
Top
Use the OneSampleKolmogorovSmirnovTest
class to test whether a sample comes from a specific distribution.
The distribution must be specified independently from the sample.
None of the parameters of the distribution can be estimated from the sample,
or the test will not be accurate.
The distribution can be
specified in two forms. It can be specified as a ContinuousDistribution
object. Alternatively, a function of one variable delegate can be specified
that evaluates the Cumulative Distribution Function of the distribution.
A Kolmogorov-Smirnov test is only valid for continuous distributions. Furthermore, the
distribution itself has to be fully specified. None of its parameters should be
estimated from the sample.
The AndersonDarlingTest and the ChiSquareGoodnessOfFitTest
are alternatives to the Kolmogorov-Smirnov test. To test whether two samples come from the
same unknown distribution, use the TwoSampleKolmogorovSmirnovTest class.
Reference