AllSplitPairings function

Variation of information for all split pairings

Calculate the variation of clustering information \insertCite Meila2007TreeDist for each possible pairing of non-trivial splits on n leaves \insertCite SmithDistTreeDist, tabulating the number of pairings with each similarity. UTF-8

AllSplitPairings(n)

Arguments

• n: Integer specifying the number of leaves in a tree.

Returns

AllSplitPairings() returns a named vector. The name of each element corresponds to a certain variation of information, in bits; the value of each element specifies the number of pairings of non-trivial splits that give rise to that variation of information. Split AB|CD is treated as distinct from CD|AB. If pairing AB|CD=CD|AB is considered equivalent to CD|AB=CD|AB (etc), then values should be divided by four.

Examples

AllSplitPairings(6)
# Treat equivalent splits as identical by dividing by four:
AllSplitPairings(6) / 4L

References

\insertAllCited

Author(s)

Martin R. Smith

(martin.smith@durham.ac.uk)