How to use recode_factor in dplyr for recoding multiple factor values?

Question

     countrycode event
1713         ESP 110mh
1009         NED    HJ
536          BLR    LJ
2882         FRA 1500m
509          EST    LJ
2449         BEL    PV
1022         EST    HJ
2530         USA    JT
2714         CUB    JT
1236         HUN  400m
238          BLR  100m
2518         USA    JT
1575         FRA 110mh
615          JPN    LJ
1144         GER    HJ
596          CAN    LJ
2477         HUN    JT
1046         GER    HJ
2501         FIN    DT
2176         KAZ    PV

I want to create a new factor vector in my data frame, eventtype, where:

Rows with 100m, 400m, 110mh, 1500m in the event variable become grouped as Runs; DT, SP, JT gets grouped as Throws; and LJ, HJ, PV, gets grouped as Jumps.

I can create a new vector value individually with something like df$eventtype <- recode_factor(df$event, `100m`="Running") works for one event, but I looked in the documentation and there isn't an easy way to convert multiple values in one function call.

Edit: of course if there is another function which serves my purposes better I will use that.

dplyr: look at case_when or in base R the ifelse statement. And you left out what 110mh should be — phiver
– phiver, Commented Apr 3, 2018 at 15:27

CJ Yetman · Accepted Answer · 2018-04-05 09:23:12Z

The ... argument of the recode_factor function can take any number of arguments...

library(dplyr)

df <- read.table(header = T, text = "
number countrycode event
1713         ESP 110mh
1009         NED    HJ
536          BLR    LJ
2882         FRA 1500m
509          EST    LJ
2449         BEL    PV
1022         EST    HJ
2530         USA    JT
2714         CUB    JT
1236         HUN  400m
238          BLR  100m
2518         USA    JT
1575         FRA 110mh
615          JPN    LJ
1144         GER    HJ
596          CAN    LJ
2477         HUN    JT
1046         GER    HJ
2501         FIN    DT
2176         KAZ    PV
")

df$eventtype <- recode_factor(df$event, `100m` = "Runs", `400m` = "Runs", 
                              `110mh` = "Runs", `1500m` = "Runs", 
                              DT = "Throws", SP = "Throws", JT = "Throws",
                              LJ = "Jumps", HJ = "Jumps", PV = "Jumps")

# or inside a mutate command
df %>% 
  mutate(eventtype = recode_factor(event, `100m` = "Runs", `400m` = "Runs", 
                                   `110mh` = "Runs", `1500m` = "Runs", 
                                   DT = "Throws", SP = "Throws", JT = "Throws",
                                   LJ = "Jumps", HJ = "Jumps", PV = "Jumps"))

Wenlong Liu · Accepted Answer · 2018-04-03 15:47:57Z

1

ifelse is what you need. Here are some sample code as you don't have a reproducible example.

countycode = c("ESP", "HUN", "KAZ")
event = c("100m", "JT", "PV")
data = as.data.frame(cbind(countycode,event))

# generate the recode groups.
runs = c("100m", "400m", "1500m")
throws = c("JT", "SP")
jumps = c("HJ", "PV")

# add another column.
data$eventtype = ifelse(data$event %in% runs, "Runs", 
                        ifelse(data$event %in% throws, "Throws",
                              ifelse(data$event %in% jumps, "Jumps",
                                     NA)))

After runs:

> data
  countycode event eventtype
1        ESP  100m      Runs
2        HUN    JT    Throws
3        KAZ    PV     Jumps

answered Apr 3, 2018 at 15:47

Wenlong Liu

4442 silver badges13 bronze badges

Collectives™ on Stack Overflow

How to use recode_factor in dplyr for recoding multiple factor values?

2 Answers 2

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related