fix an incomplete difference matrix by guessing from differences between locations
close to the ones whose values are missing
Suppose you got a difference matrix, created by the program
leven, that has a few missing values
(
NA's). Not all programs can handle a difference matrix with
missing values, for example:
mds,
cluster. This program,
diffix, makes reasonable guesses to fix an incomplete difference
matrix.
Suppose there is no difference value available for locations A
and B. This value is substituted by the average of the following 8
values:
- The differences between A and the 4 locations closest to
B for which a difference with A is available.
- The differences between B and the 4 locations closest to
A for which a difference with B is available.
To determine what locations are closest to the ones involved in
missing differences, it must me determined how the coordinates must be
interpreted.
Use option -L if the coordinates are given in
longitude and latitude.
Use option -a if the coordinates are given in some user-defined
rectangular grid, and specify the aspect ration between X- and
Y-coordinates. This must be a value of 1.0 or greater, meaning a distance
of 1X can represent a smaller geographical distance than a distance of 1Y.