# install.packages('rio')
library(rio)
datakab = import('https://raw.githubusercontent.com/msaidf/statek/master/content/indo-dapoer_data.csv')

datakab

nrow(datakab)

head(datakab, n = 3)

tail(datakab)

tail(datakab[1:12012,], 2)

library(magrittr)
datakab[-12013:-12017,] %>% tail(3)

which(datakab$`Series Code` != "") %>% datakab[.,] %>% tail(2)

dplyr::filter(datakab, `Series Code` != "")  %>% tail(2)

datakab = dplyr::filter(datakab, `Series Code` != "") 

c(1, 3:5, ncol(datakab)) %>% datakab[, .] %>% tail(4)

datakab[,c('Region Code', 'Series Name')] %>% tail

library(dplyr)

select(datakab, 'Region Code':'Series Code', -'Series Name', 
       region_name = 'Region Name', contains('YR'), -contains('YR201')) %>% tail(4)

rename(datakab, region_name = 'Region Name', series_code = 'Series Code') [100:102,]

names(datakab)[1:3] = names(datakab)[1:3] %>% toupper
names(datakab) %>% t

datakab = janitor::clean_names(datakab)
names(datakab) %>% t

datakab %<>% rename_at(vars(starts_with('x20')), 
                       funs(stringr::str_replace(., 'x20[0-9][0-9]_', ''))) 
datakab %>% names %>% t

data1 = datakab %>% select(region_name:series_code, contains('201'))
names(data1) %>% t

data2 = datakab %>% select(-contains('201'))
names(data2) %>% t

merge(data1, data2, by = c('region_code', 'series_code')) %>% tail(3)

dplyr::inner_join(data1, data2) %>% tail(3)

datakab[2,5]

(datakab[2,5] = NA)

datakab %>% select(yr2000, yr2001) %>% 
            summarytools::dfSummary(graph.col = F, na.col = F) %>% 
            print

datakab %>% mutate(yr2000 = ifelse(yr2000 == '..', NA, yr2000),
                   yr2001 = ifelse(yr2001 == '..', NA, yr2001)) %>% 
            select(yr2000:yr2002) %>% 
            summarytools::dfSummary(graph.col = F, na.col = F) %>% 
            print

datakab %<>% mutate_at(vars(starts_with("yr20")), funs(ifelse(. == '..', NA, .))) 

datakab %<>% mutate_at(vars(starts_with("yr20")), as.numeric)

datakab %>% select(starts_with("yr201")) %>%
            summarytools::dfSummary(graph.col = F, na.col = F) %>% 
            print

datakab %<>% gather(year, val, starts_with('yr'))  
datakab %>% head(2)

datakab %<>% mutate(year = str_replace(year, 'yr', '') %>% 
                          as.integer) %>% 
            select(-series_code) %>% 
            spread(key = series_name, value = val)
datakab %>% head(2)

datakab %<>% janitor::clean_names()
names(datakab) %<>% str_replace_all('_in_.+$', '') %>% t
datakab %>% names %>% t

save.image(file = ".RData")
save(datakab, file = "datakab.rda")

rm(data1, data2)
ls()

rm(list = ls())
ls()

load('datakab.rda')
ls()

saveRDS(datakab, file = "datakab.rds")
indodapoer = readRDS('datakab.rds')
ls()