I can't find a simple library function for k-centers clustering using R, whereas I could for k-means (kmeans()) and hierarchical clustering (hclust()).
Is there a library function for simple greedy k-centers clustering using R as depicted in this post
If not - as I am new to R - how would one go about implementing it (I understand the logic - just not how to actually write it in R code).
Try kmeans with method = "centers".
If this is not what you are looking for, then CRAN has a cluster Task View with dozens of packages at http://cran.r-project.org/web/views/Cluster.html.
From what is described in this blog post, this seems similar to one of the many seeding strategies used for k-means. I wouldn't really call it a clustering method yet, but a pre-clustering or something like that.
Maybe you should lookt at the flexclust package of R, I believe it has some k-means variations and initializations, and maybe it has this variant as one initialization option. Or it might be on http://cran.r-project.org/web/views/Cluster.html
Note that always choosing the object that is furthest away is prone to choose outliers as cluster centers! Have a look at e.g. k-means++ which is based on a similar idea, but somewhat more clever (plus, it better supports randomization, so you can try multiple different initializations). Or you could choose the object that is closest to the (2k-1)/(2k) quantile, which probably is a better guess for a good cluster center.
