Because some information is common to both splits (SplitMutualInformation
),
the joint information of two splits will be less than the sum of the
information of the splits taken separately -- unless the splits are
contradictory.
Split Y1 is defined as dividing taxa into the two sets A1 and B1,
and Y2=A2:B2.
Consider partitions that divide eight terminals, labelled A to H.
Bipartition 1: |
ABCD:EFGH |
A1 = ABCD |
B1 = EFGH |
This can be represented by an association matrix:
The joint information is given by
JointInformation(3, 1, 0, 4)
.