I have a dataset contains a column like this:
dput(df[1:20, 8]) %>% structure()
[1] "TASBUDAK-EREGLI (KONYA)"
[2] "HANCAGIZ-SAMANDAG (HATAY)"
[3] "SOGUT-DOGANSEHIR (MALATYA)"
[4] "KUCUKLU-DOGANSEHIR (MALATYA)"
[5] "KEMALPASA-GOKSUN (KAHRAMANMARAS)"
[6] "ULUKOY-(MALATYA)"
[7] "KAZANLIPINAR-(KAHRAMANMARAS)"
[8] "ULUBAHCE-PAZARCIK (KAHRAMANMARAS)"
[9] "EMIRLER-NURDAGI (GAZIANTEP)"
[10] "CELIKKOY-GOLBASI (ADIYAMAN)"
[11] "KAZANDERE-GOKSUN (KAHRAMANMARAS)"
[12] "BESCI-EMIRGAZI (KONYA)"
[13] "ORDEKDEDE-PAZARCIK (KAHRAMANMARAS)"
[14] "CAVUSLU-DOGANSEHIR (MALATYA)"
[15] "KULLAR-NURHAK (KAHRAMANMARAS)"
[16] "IZCI-SURUC (SANLIURFA)"
[17] "YAZIKOY-AFSIN (KAHRAMANMARAS)"
[18] "DEMIRCI-EMIRGAZI (KONYA)"
[19] "ORTULU-LICE (DIYARBAKIR)"
[20] "KAMISTAS-(MALATYA)"
I want to separate this column into three columns with the code below:
df %>%
separate(Region,
into = c("Suburb", "Town", "City"))
and the result is that:
Suburb | Town | City |
---|---|---|
KAZANDERE | GÖKSUN | KAHRAMANMARAS |
DOGU AKDENIZ | NA | NA |
ISIKLI | YESILYURT | MALATYA |
COBANTASI | BINGOL | NA |
GOKSUN | KAHRAMANMARAS | NA |
YAGMURLU | GOKSUN | KAHRAMANMARAS |
But i want to split this columns with a right justified order like that:
Suburb | Town | City |
---|---|---|
KAZANDERE | GÖKSUN | KAHRAMANMARAS |
NA | NA | DOGU AKDENIZ |
ISIKLI | YESILYURT | MALATYA |
NA | COBANTASI | BINGOL |
NA | GOKSUN | KAHRAMANMARAS |
YAGMURLU | GOKSUN | KAHRAMANMARAS |
How can I do that?
df %>%
separate(Region, into = c("Suburb", "Town", "City"), fill = 'left')
should do the trick