R/clu_means.R

In clustrd: Methods for Joint Dimension Reduction and Clustering

clu_means<-function(x, id, disp=TRUE, center=TRUE, scale=TRUE){
  
  clu = NULL
  #funs = NULL
  
  x = data.frame(scale(as.matrix(x), center = center, scale = scale), stringsAsFactors = TRUE)
  
  p=ncol(x)
  gm=apply(x,2,mean)
  
  id<-factor(id)
  csize=as.vector(table(id)/sum(table(id)))
  
  x$clu=id
  clum=(x %>% group_by(clu) %>% summarise_all(list(mean)))
  am=rbind(clum[,-1],gm)
  bm=data.frame(t(am),stringsAsFactors = TRUE)
  names(bm)=c(paste("C",1:nrow(clum),sep=""),"all")
  bm$names=row.names(bm)
  
  par_bm=data.frame(t(bm[-ncol(bm)]),stringsAsFactors = TRUE)
  gnam=paste(names(bm)[-ncol(bm)]," (",round(csize*100,digits=1),"%",")",sep="")
  #  cnm=paste(cnames,": ",round(csize*100,2),"%",sep="")
  
  gnam[length(gnam)] = "all"
  par_bm$clusters=gnam
  par_bm$csize=c(csize,1/length(csize))
  
  gg_color_hue <- function(n) {
    hues = seq(15, 375, length = n + 1)
    hcl(h = hues, l = 65, c = 100)[1:n]
  }
  
  mypal=gg_color_hue(length(csize))
  mypal=c("black",mypal)
  
  # if (scale == T) {
  #    pco=ggparcoord(par_bm[1:(dim(par_bm)[1]-1),],columns=1:p,groupColumn=p+1,scale="globalminmax",mapping = ggplot2::aes(size = 5*csize)) 
  #  } else {
  pco=ggparcoord(par_bm,,columns=1:p,groupColumn=p+1,scale="globalminmax",mapping = ggplot2::aes(size = 3*csize))
  #  }
  pco=pco+scale_size_identity()
  pco=pco+scale_colour_manual(values=mypal)
  
  #  if (scale == T) {
  #    pco=pco+geom_vline(xintercept=1:p,alpha=.5) + xlab("variables") + ylab("z-score") 
  #  } else {
  pco=pco+geom_vline(xintercept=1:p,alpha=.1) + xlab("") + ylab("mean") + theme_classic() + theme(axis.text.x = element_text(hjust=1, angle=45))
  #  }
  
  return(pco)
  
}