Ifelse for Multiple Columns in DataFrame-CodePudding

I have a dataset like the following:

ID	Winter	Spring	Summer	Fall
1	high	NA	high	low
2	low	high	NA	low
3	low	NA	NA	low
4	low	high	NA	low

I would like to add a calculated column so that if any of the winter, spring, summer, and fall columns contains "high," then "1" is added to that row as shown below. Otherwise it will contain 0.

ID	Winter	Spring	Summer	Fall	calculated_column
1	high	NA	high	low	1
2	low	high	NA	low	1
3	low	NA	NA	low	0
4	low	high	NA	low	1

So far I have something like this, I know it's incorrect. I'm not sure how to specify multiple columns rather than just one:

df$calculated_column <- ifelse(c(2:5)=="High",1,0)

CodePudding user response：

We may use if_any

library(dplyr)
df1 <- df1 %>%
     mutate(calculated_column =  (if_any(-ID, ~ . %in% 'high')))

-output

df1
 ID Winter Spring Summer Fall calculated_column
1  1   high   <NA>   high  low                 1
2  2    low   high   <NA>  low                 1
3  3    low   <NA>   <NA>  low                 0
4  4    low   high   <NA>  low                 1

Or if we want to use base R, create the logical condition with rowSums on a logical matrix

df1$calculated_column <-   (rowSums(df1[-1] == "high", na.rm = TRUE) > 0)

data

df1 <- structure(list(ID = 1:4, Winter = c("high", "low", "low", "low"
), Spring = c(NA, "high", NA, "high"), Summer = c("high", NA, 
NA, NA), Fall = c("low", "low", "low", "low")), 
class = "data.frame", row.names = c(NA, 
-4L))

CodePudding user response：

You could also do:

df1$calculated_column =  grepl('high', do.call(paste, df1))
df1
  ID Winter Spring Summer Fall calculated_column
1  1   high   <NA>   high  low                 1
2  2    low   high   <NA>  low                 1
3  3    low   <NA>   <NA>  low                 0
4  4    low   high   <NA>  low                 1

CodePudding user response：

Here is a base R solution:

calculated_column = (apply(df1,1,function(x) sum(grepl("high",x)))>0)*1

cbind(df1, calculated_column)

  ID Winter Spring Summer Fall calculated_column
1  1   high   <NA>   high  low                 1
2  2    low   high   <NA>  low                 1
3  3    low   <NA>   <NA>  low                 0
4  4    low   high   <NA>  low                 1