# see kpeaks documentation at https://cran.r-project.org/web/packages/kpeaks/kpeaks.pdf
# publication at https://www.researchgate.net/publication/331258718_kpeaks_An_R_Package_for_Quick_Selection_of_K_for_Cluster_Analysis
# you'll need to install jsonlite and kpeaks packages into Vertica R UDx as shown at https://www.vertica.com/docs/9.2.x/HTML/Content/Authoring/ExtendingVertica/R/RPackages.htm
library("jsonlite")
library("kpeaks")

KPeaks_User <- function(input.data.frame, parameters.data.frame) {
  clusters <- findk(input.data.frame, binrule="bc", rcs=FALSE)
  dfSize <- nrow(input.data.frame)
  vector <- character()
  for (i in 0:dfSize)
    vector[i] <- toJSON(clusters)
  output <- data.frame(vector)
  return(output)
}

KPeaks_UserFactory <- function() {
  list(name    = KPeaks_User,
       udxtype = c("scalar"),
       # Since this is a polymorphic function the intype must be any
       intype  = c("any"),
       outtype = c("varchar(255)"),
       parametertypecallback=KPeaksParameters)
}

# I will likely extend this to support optional parameters to kpeaks like binrule
KPeaksParameters <- function() {
  parameters <- list(datatype = c("varchar", "int"),
                     length   = c("NA", "NA"),
                     scale    = c("NA", "NA"),
                     name     = c("binrule", "nstart"))
  return(parameters)
}