# Collocated Option (Unique Neighborhood)

This is an interesting case for:

- estimating a Target
- in a multivariate case (say with N variables)
- based on the information in the input Db (note that all 'N' variable(s) do not have to be known in the 'heterotopic' case)
- when the Target contains information on some of the 'N' variables (these are the **collocated** variables) 

When working in **Unique** Neighborhood, the relevant information (from the input data base) do not change for any of the targets. But the Collocated information changes at each target.

Hence the interest of benefiting of the inversion of the covariance matrix (restricted to the information of the Input data base).