Replace values in a column with specific row value from same column using loop

问题

I have data obtained from a survey that lists the recipient's name and whether or not they selected a specific county in the state. The survey structure outputs an off for any county not selected and an for the selected county. The state has about 100 counties so there end up being a lot of columns that really correspond to the same question. What I am looking to do is replace any cells with on with the county name and any cells with off with a blank. From there I can basically unite many columns into one without much difficulty. Below I have recreated a brief example data set

name <- c("Recipient", "AB", "BC", "DF", "EF", "WE")
Q1 <- c("County1", "Off", "On", "On", "Off", "Off")
Q2 <- c("County2", "On", "Off", "Off", "Off", "Off")
Q3 <- c("County3", "Off", "Off", "Off", "On", "On")
dt <- data.frame(name, Q1, Q2, Q3)
> dt
       name       Q1           Q2           Q3
1 Recipient  County1      County2      County3
2        AB      Off           On          Off
3        BC       On          Off          Off
4        DF       On          Off          Off
5        EF      Off          Off           On
6        WE      Off          Off           On

I am looking for a desired output of

       name       Q1           Q2           Q3
1 Recipient  County1      County2      County3
1        AB               County2     
2        BC County1           
3        DF County1            
4        EF                             County3
5        WE                             County3

I am not sure how to go about this and designate that the first row be used to fill cells.

Thanks for any help.

回答1:

We create a logical vector and assign the first row values based on the logical vector

i1 <- dt[-1] == 'On'
dt[-1][i1] <- unlist(dt[1, -1])[col(dt[-1])][i1]
dt[-1][!i1] <- ""
dt
#       name       Q1       Q2       Q3
#1 Recipient                           
#2        AB          County 2         
#3        BC County 1                  
#4        DF County 1                  
#5        EF                   County 3
#6        WE                   County 3

Or with dplyr

library(dplyr)
dt %>% 
  mutate_at(vars(starts_with('Q')), ~ case_when(. == 'On' ~first(.), TRUE ~ ''))

data

dt <- data.frame(name, Q1, Q2, Q3, stringsAsFactors = FALSE)

来源：https://stackoverflow.com/questions/59669075/replace-values-in-a-column-with-specific-row-value-from-same-column-using-loop

标签

loops

replace

dplyr

mutate