Computing Distance Matrix based on Point Vectors (or coordinate matrix) and its reverse: Computing Point Vectors based on Distance Matrix
There are 2 functions:
- computes the distance matrix D given vector of points V. (this is the easy part)
- given distance matrix D, derives the relative points from the centroid (the reverse)
Note: (my observation only)
- DM = getDM(V)
- V' = getXY(DM)
Here, V and V' may not have the same points. V is V' when V is centered as the centroid, and probably rotated. in other words, V' = HV where H is a transform matrix (similarity, i.e, by location and rotation only)
Sources:
- Teknomo's Multi Dimensional Scaling, Pdf file
- https://en.wikipedia.org/wiki/Multidimensional_scaling