CODEWITHSUNDEEP
rif-statementindexing
user15141497
user15141497

Reputation: 73

Defining element based on condition in R

Say that I have a dataframe (df) that looks like this:

stim <- c("a", "k", "y", "j", "t", "Stop", "f", "l", "b", "a", "c", "Stop")
df <- data.frame(stim)

I would like to create another column 'YesorNo' based on the value of 'stim'. My desired output is this:

stim  YesorNo
 a      yes
 k      yes
 y      yes
 j      yes
 t       No
 Stop   N/A
 f      yes
 l      yes
 b      yes
 a      yes
 c       No
 Stop   N/A
 ...    ...

I would like the value of 'YesorNo' to be N/A if stim is "stop". I can use a simple ifelse statement for that (df$YesorNo <- ifelse(df$stim=="Stop", "N/A", "yes") But I would also like the value that comes BEFORE "stop" (letters t and c in my example) to correspond to "no" instead of "yes". How would I define that?

Upvotes: 2

Views: 41

Answers (3)

Ronak Shah
Ronak Shah

Reputation: 388817

You can use lead to look for the next value and case_when to list the conditons.

library(dplyr)

df %>%
  mutate(YesorNo = case_when(stim == 'Stop' ~ NA_character_, 
                             lead(stim) == 'Stop' ~ "No", 
                             TRUE ~ 'Yes'))

#   stim YesorNo
#1     a     Yes
#2     k     Yes
#3     y     Yes
#4     j     Yes
#5     t      No
#6  Stop    <NA>
#7     f     Yes
#8     l     Yes
#9     b     Yes
#10    a     Yes
#11    c      No
#12 Stop    <NA>

Similarly in data.table you can use fcase.

setDT(df)[, YesorNo := fcase(stim == 'Stop', NA_character_, 
                             shift(stim, type = 'lead') == 'Stop', "No", 
                             default = "Yes")]
df

Upvotes: 3

Kra.P
Kra.P

Reputation: 15123

I'm sorry that I don't know how to call next or past value in dplyr

df<- df%>%
  mutate(YesorNo = ifelse(stim == "Stop", "N/A", "yes"))
for (i in 1:dim(df)[1]){
  if (df$YesorNo[i] == "N/A") {
    if(i == 1){
    } else {
      df$YesorNo[i-1] = "no"
    }
  }
}

   stim YesorNo
1     a     yes
2     k     yes
3     y     yes
4     j     yes
5     t      no
6  Stop     N/A
7     f     yes
8     l     yes
9     b     yes
10    a     yes
11    c      no
12 Stop     N/A

Upvotes: 2

Vin&#237;cius F&#233;lix
Vin&#237;cius F&#233;lix

Reputation: 8811

Here is a tidyverse solution using case_when + operator %in%

Code

library(tidyverse)
    
df %>% 
  mutate(
    YesorNo = case_when(
      stim == "Stop" ~ NA_character_,
      stim %in% c("t","c") ~ "no",
      #other conditions
      TRUE ~ "yes"
    )
  )

Result

   stim YesorNo
1     a     yes
2     k     yes
3     y     yes
4     j     yes
5     t      no
6  Stop    <NA>
7     f     yes
8     l     yes
9     b     yes
10    a     yes
11    c      no
12 Stop    <NA>

Upvotes: 2

Related Questions