Interface MCDETest<R extends MCDETest.RankStruct>
-
- Type Parameters:
R
- RankStruct or Subclass of RankStruct
- All Known Implementing Classes:
MWPTest
public interface MCDETest<R extends MCDETest.RankStruct>
Interface for statistical tests forMCDEDependence
. SeeMWPTest
for an example. Implementation should contain an appropriate index structure to efficiently compute statistical test and the test itself.- Since:
- 0.8.0
- Author:
- Alan Mazankiewicz, Edouard Fouché
-
-
Nested Class Summary
Nested Classes Modifier and Type Interface Description static class
MCDETest.RankStruct
Structure to hold return values in index creation for MCDEDependence
-
Method Summary
All Methods Instance Methods Abstract Methods Modifier and Type Method Description <A> R
correctedRanks(NumberArrayAdapter<?,A> adapter, A data, int len)
Compute the corrected rank index.double
statisticalTest(int start, int width, boolean[] slice, R correctedRanks)
Subclass must implement the computation of the statistical test, based on the slicing scheme of MCDEDependence.
-
-
-
Method Detail
-
correctedRanks
<A> R correctedRanks(NumberArrayAdapter<?,A> adapter, A data, int len)
Compute the corrected rank index.- Type Parameters:
A
- Input array type- Parameters:
adapter
- Array-like adapterdata
- Data objectlen
- Length- Returns:
- corrected ranks
-
statisticalTest
double statisticalTest(int start, int width, boolean[] slice, R correctedRanks)
Subclass must implement the computation of the statistical test, based on the slicing scheme of MCDEDependence.- Parameters:
start
- Starting index value for statistical testwidth
- Width of the slice (endindex = start + width)slice
- An array of boolean resulting from a random slicecorrectedRanks
- the precomputed index for the reference dimension- Returns:
- a 1 - p-value
-
-