tabla_con_clave <- readRDS("censos/censo_personas_con_clave_17")
tabla_con_clave_u <- filter(tabla_con_clave, tabla_con_clave$AREA == 1)
b <- tabla_con_clave_u$COMUNA
c <- tabla_con_clave_u$P16A_GRUPO
cross_tab =  xtabs( ~ unlist(b) + unlist(c))
tabla <- as.data.frame(cross_tab)
d <-tabla[!(tabla$Freq == 0),]
d$anio <- "2017"
d_t <- filter(d,d$unlist.c. == 1)
for(i in 2:10){
  d_i <- filter(d,d$unlist.c. == i)
  d_t = merge( x = d_t, y = d_i, by = "unlist.b.", all.x = TRUE)
}

codigos <- d_t$unlist.b.
rango <- seq(1:nrow(d_t))
cadena <- paste("0",codigos[rango], sep = "")
cadena <- substr(cadena,(nchar(cadena)[rango])-(4),6)
codigos <- as.data.frame(codigos)
cadena <- as.data.frame(cadena)
comuna_corr <- cbind(d_t,cadena)
comuna_corr <- comuna_corr[,-c(1),drop=FALSE] 
names(comuna_corr)[ncol(comuna_corr)] <- "código"

quitar <- seq(3,(ncol(comuna_corr)-1),3)
comuna_corr <- comuna_corr[,-c(quitar),drop=FALSE]

names(comuna_corr)[2] <- "Mapuche"
names(comuna_corr)[4] <- "Aymara"
names(comuna_corr)[6] <- "Rapa Nui"
names(comuna_corr)[8] <- "Lican Antai"
names(comuna_corr)[10] <- "Quechua"
names(comuna_corr)[12] <- "Colla"
names(comuna_corr)[14] <- "Diaguita"
names(comuna_corr)[16] <- "Kawésqar"
names(comuna_corr)[18] <- "Yagán o Yamana"
names(comuna_corr)[20] <- "Otro"

quitar <- seq(1,(ncol(comuna_corr)-1),2)
comuna_corr <- comuna_corr[,-c(quitar),drop=FALSE]
ingresos_expandidos_2017 <- readRDS("Ingresos_expandidos_urbano_17.rds")
df_2017_2 = merge( x = comuna_corr, y = ingresos_expandidos_2017, by = "código", all.x = TRUE)

union_final_urb <- df_2017_2[,-c(1,(ncol(df_2017_2)-1))]
write_xlsx(union_final_urb, "P16A_GRUPO_urbano.xlsx")

data_sum <- summary(union_final_urb)
kbl(head(data_sum)) %>%
  kable_styling(bootstrap_options = c("striped", "hover")) %>%
  kable_paper() %>%
  scroll_box(width = "100%", height = "500px")

library(plotly)

df_2017_fig <- df_2017_2[,-c((ncol(df_2017_2)-1))]

fig <- plot_ly(df_2017_fig, x = df_2017_fig$código, y = df_2017_fig[,2]
, name = colnames(df_2017_fig[2]), type = 'scatter', mode = 'lines',
                 width=7000, height=400) 

grafica_fn <- function(g){
  fig <<- fig %>% add_trace(y = ~df_2017_fig[,g]
, name = colnames(df_2017_fig[g]), mode = 'lines',
                 width=7000, height=400) 
}
for (g in 3:(ncol(df_2017_2)-1)) {
grafica_fn(g)

}

fig <- fig %>% layout(autosize = T )
fig 

chart.Correlation(union_final_urb, histogram=TRUE, method = c( "pearson"), pch=20)

chart.Correlation(union_final_urb, histogram=TRUE, method = c( "spearman"), pch=20)

df_2017_2f <- filter(union_final_urb, union_final_urb$ingresos_expandidos != 'is.na')
III <- seq(1,(ncol(df_2017_2f)-1),1)
my_data <- df_2017_2f[, c(III)]
tabla <- cor(x=my_data, y=df_2017_2f$ingresos_expandidos, method=c("kendall"), use = "pairwise")
tabla <- as.data.frame(tabla) 
colnames(tabla) <- "Correlación" 
saveRDS(tabla,"tablas_de_corr/C_P16_GRUPO_RU.rds")
tabla %>%
  rownames_to_column("ueblo indígena u originario (grupo)") %>%  
  mutate(Correlación = cell_spec(Correlación, background=ifelse(Correlación == max(Correlación), "#fc0303", "#5cb81f"))) %>% 
  kbl(booktabs = T, linesep = "", escape=FALSE) %>% 
    kable_paper(full_width = F) %>%
    column_spec(1, color = "black")%>%
    column_spec(2, color = "white")

chart.Correlation(union_final_urb, histogram=TRUE, method = c( "kendall"), pch=20)

chart.Correlation(union_final_urb, histogram=TRUE, method = c( "pearson"), pch=20)

chart.Correlation(union_final_urb, histogram=TRUE, method = c( "spearman"), pch=20)