Use R dist() function to compute similarity across spatial units

[mbacou]

Would be an improvement to hcapi::similar() https://ifpri.github.io/sda-hcapi3/reference/similar.html choosing one of:

• Euclidean - The most common measurement of distance. sqrt(sum((x - y)^2))
• Maximum - Maximum difference between elements of two vectors.
• Manhattan - Sum of difference in each elements of two vectors.
• Canberra - Manhattan distance weighted by the elements. sum(|x_i - y_i| / |x_i + y_i|)
• Binary - non-zero elements are regarded as 1 and zero elements are regarded as 0. The distance is the number of elements only one is 1 divided by numbers of them at least one is 1.
• Minkowski - This is generalized case of euclidean and manhattan. If p=2, it becomes euclidean and if p=1, it becomes manhattan. sum((x - y)^p)^(1/p)