rm(list = ls()) # limpio memoria
linkGoogle="https://docs.google.com/spreadsheets/d/e/2PACX-1vQOKTNeLrYhFklfcGWO9-xzVuYFuEGv-LlomX5ou3dXxNye6XmSYj6pOoYZgfThB2p2vUEn29jLFe1F/pub?gid=1366853308&single=true&output=csv"
allDFs=read.csv(linkGoogle)
allDFs$Electricidad_pct=allDFs$elec1_Sí/allDFs$elec3_Total
allDFs$CastilloKeiko_ratio=allDFs$Castillo/allDFs$Keiko
allDFs$fallecido_x10000POS=1000*allDFs$countFallecidos/allDFs$countPositivos
keptForCluster=c('key','Electricidad_pct','CastilloKeiko_ratio','fallecido_x10000POS')
allDFs_small=allDFs[,keptForCluster]
allDFs_small=allDFs_small[!allDFs_small$key=='LIMA+LIMA',]
allDFs_small_norm=allDFs_small
allDFs_small_norm[,-1]=BBmisc::normalize(allDFs_small[,-1],method='standardize')
dataClus=allDFs_small_norm[,-1]
row.names(dataClus)=allDFs_small_norm$key
dataClus=dataClus[complete.cases(dataClus),]
g.dist = cluster::daisy(dataClus, metric="gower")
factoextra::fviz_nbclust(dataClus, factoextra::hcut,diss=g.dist,method = "gap_stat",k.max = 10,verbose = F,hc_func = "agnes")
factoextra::fviz_nbclust(dataClus, factoextra::hcut,diss=g.dist,method = "gap_stat",k.max = 10,verbose = F,hc_func = "diana")
