Replace values in r using case_when

Question

I'm trying to replicate some Stata code in r. In my df, there is a variable "time" and one "exposure" (both numeric, so with values like 1,2,3 etc.

This is what the original Stata code looks like:

replace time = (time - 31) if exposure == 0

And this is what I tried:

recoded_df <- recoded_df %>% mutate(time = case_when(exposure == 0 ~ (time - 31))

I get the error that there is an "unexpected symbol" in the code. But I have no idea what I am doing wrong.

Could you please share a reproducible piece of your data by means of dput(head(data))? It's essential if you are to have any chances of getting relevant solutions. — Anoushiravan R
– Anoushiravan R, Commented May 25, 2021 at 15:47
Without sample data to test, your code looks right except that you are missing a ) at the end. You probably also want to set a default of "no changes", mutate(time = case_when(exposure == 0 ~ (time - 31), TRUE ~ time)) — Gregor Thomas
– Gregor Thomas, Commented May 25, 2021 at 15:52
In addition to sample data, I can't figure out how to get that error from this code: without the last right-paren, R just waits for me to close out the expression, no error. My guess is that you have something else, perhaps mutate(time = case_when(...) time2 = foo(.)), where there needs to be a comma (just a guess, though). — r2evans
– r2evans, Commented May 25, 2021 at 15:55
@GregorThomas, I tend to prefer that methodology as well, explicitly adding a default value, but case_when does default to NA which might be what the OP is intending. (idk) — r2evans
– r2evans, Commented May 25, 2021 at 15:56
And @r2evans my assumption is the Stata code shared would not make the non-transformed values NA. — Gregor Thomas
– Gregor Thomas, Commented May 25, 2021 at 15:57

Zulkifli · Accepted Answer · 2021-05-25 16:46:29Z

2

library(dplyr)

df <- data.frame(
  exposure = c(0, 1),
  time = c(81, 20)
)

df %>%
  mutate(
    time = case_when(exposure == 0 ~ time - 30,
                     TRUE ~ time)
  )

  exposure time
1        0   51
2        1   20

answered May 25, 2021 at 16:46

Zulkifli

943 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

akrun · Accepted Answer · 2021-05-25 17:23:39Z

1

Using data.table, fifelse

library(data.table)
setDT(df)[, time := fifelse(!exposure, time - 30,  time)]

answered May 25, 2021 at 17:23

akrun

891k38 gold badges590 silver badges700 bronze badges

Comments

TarJae · Accepted Answer · 2021-05-25 17:13:27Z

0

Here are two possibilities how you can achieve your task with a fake dataframe:

# fake dataframe:
df <- tribble(
  ~time, ~exposure,
  100, 0,
 70, 1, 
  90, 0, 
  60, 1, 
  50, 0
)

# 1. with mutate and ifelse
library(dplyr)
df %>% 
  mutate(time = ifelse(exposure == 0, time-31, time))

# 2. with base R and ifelse
df$time <- ifelse(df$exposure == 0, df$time - 31, df$time)

answered May 25, 2021 at 17:13

TarJae

80.2k6 gold badges30 silver badges94 bronze badges

Comments

Tim Biegeleisen · Accepted Answer · 2021-05-25 15:50:54Z

0

If I understand correctly, ifelse from base R should cover this:

recoded_df$time = ifelse(recoded_df$exposure == 0,
                         recoded_df$time - 31, recoded_df$time)

answered May 25, 2021 at 15:50

Tim Biegeleisen

526k32 gold badges324 silver badges399 bronze badges

Collectives™ on Stack Overflow

Replace values in r using case_when

4 Answers 4

Comments

Comments

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

4 Answers 4

Comments

Comments

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related