问题
I'm working with a fairly large data set (100k rows) and want to replicate the Excel Index Match function in R Studio.
I'm looking for a way to create a new column that will pull a value from an existing column ("1995_Number"), if 3 values from three different columns from one year match three values from three columns from another year - independent of the rows, and create a new column ("1994_Number").
Dataframe as example:
dat <- data.frame(`1994_Address` = c("1234 Road", "123 Road", "321 Road"),
`1994_ZipCode` = c(99999, 99999, 11111),
`1994_Bank Name` = c("JPM", "JPM", "WF"),
`1995_Address` = c("123 Road", "1234 Road", "321 Road"),
`1995_ZipCode` = c(99999, 99999, 11111),
`1995_Bank Name` = c("JPM", "JPM", "WF"),
`1995_Number` = c(1, 2, 3), check.names = F, stringsAsFactors = F)
The newly created column 1994_Number should say (2, 1, 3)
回答1:
A possible solution would include the match
function from base
. Toghether with dplyr
the following works:
library(dplyr)
dat <- data.frame(`1994_Adress` = c("1234 Road", "123 Road", "321 Road"),
`1994_ZipCode` = c(99999, 99999, 11111),
`1994_Bank Name` = c("JPM", "JPM", "WF"),
`1995_Adress` = c("123 Road", "1234 Road", "321 Road"),
`1995_ZipCode` = c(99999, 99999, 11111),
`1995_Bank Name` = c("JPM", "JPM", "WF"),
`1995_Number` = c(1, 2, 3), check.names = F, stringsAsFactors = F)
dat %>%
mutate(`1994_Number` = ifelse(`1994_Adress` %in% `1995_Adress` &
`1994_ZipCode` %in% `1995_ZipCode` &
`1994_Bank Name` %in% `1995_Bank Name`,
dat[match(dat$`1994_Adress`, dat$`1995_Adress`), "1995_Number"], NA))
# 1994_Adress 1994_ZipCode 1994_Bank Name 1995_Adress 1995_ZipCode 1995_Bank Name 1995_Number 1994_Number
# 1 1234 Road 99999 JPM 123 Road 99999 JPM 1 2
# 2 123 Road 99999 JPM 1234 Road 99999 JPM 2 1
# 3 321 Road 11111 WF 321 Road 11111 WF 3 3
来源:https://stackoverflow.com/questions/61795811/index-match-in-r-studio-multiple-columns-across-rows