Shimbel Distance Matrix (D-Matrix)
The Shimbel Distance Matrix (or D-Matrix) holds the shortest paths
between the nodes of a network, which are always equal or lesser to
the diameter. To construct this matrix, C matrices of Nth order are
built until the diameter of the network is reached. Each C matrix is
converted in a corresponding D matrix. In this case, two C matrices,
C1(connectivity matrix) and C2 (two-linkages paths; C1*C1) are built
since the diameter is 2.
- The first order Shimbel Matrix (D1) is a simple adaptation
of C1, where all the direct links are kept. A value of 0 is assigned
for all the cii cells since the shortest path between a node and
itself is always 0. Cells that have a value of 0 in the C1 matrix
(outside cii cells) remain unfilled on the D1 matrix.
- The second order Shimbel Matrix (D2) is built from the
first order matrix D1 but only from its unfilled cells. A value
of 2 is assigned for each cells on the D2 matrix that have a value
greater than 0 on the C2 matrix, but if a value of 1 already exists
(D1 matrix), this value is kept. This means that on the D2 matrix
of the above figure, only the values of the yellow cells have been
changed to 2. Since the diameter of this network is 2, the D2
matrix is the Shimbel distance matrix.
- Nth order Shimbel Matrix (DN). For a network having a
diameter of 3, a D3 matrix would have to be built from a C3 matrix
(C1*C2) because at least 1 cell would have remained empty in the
D2 matrix. Repeat the construction of Nth order Shimbel matrices
until the diameter is reached.
- The Shimbel Matrix (D). The order of the Shimbel distance
matrix that corresponds to the diameter is the D matrix. The summation
of rows or columns represents the Shimbel distance for each node.
In the D matrix of the above example, node C is having the least
summation of shortest paths (4) and is thus the most accessible,
followed by node A (5), nodes B and D (6) and node E (7). The total
summation of minimal paths is 28.