Custom distance metric in R for Agglomerative clustering

Question

Is it possible for us to define a custom distance function for agglomerative clustering in R? Also, I would like to prevent two clusters from being merged when a certain condition is not satisfied. Is this possible in R?

Also, is it possible to do all this when R is put on Hadoop?

Answer 1

in R, for agglomerative hierarchical clustering, there are several distance measure you could use: Euclidean, Absolute Distance (Manhattan)or Absolute Distance (Manhattan) depend on your data type.Try to look at this:http://127.0.0.1:29417/library/cluster/html/agnes.html or just go to R website (http://www.r-project.org/) and look at the manual. It may help.