CODEWITHSUNDEEP
r
doncarlos
doncarlos

Reputation: 401

if else with multiple conditions and NA

Using the example provided below I need to use a if else statement with the following conditions:

  1. if df$flag = 0 than use df$corrected
  2. if df$flag = 1 use df$original if it is larger than corrected otherwise use df$corrected
  3. if df$flag = NA use df$original

Sample data:

df <- data.frame(
  original =  c(20,40,20,2,20,10,12),
  corrected = c(25,43,22,1,25,7,NA),
  flag = c(0,0,0,1,1,1,NA))

The output should be appended to the df (df$final) and should, for this example output this:

df$final = c(25,43,22,2,25,10,12)

Thanks for any suggestions.

Upvotes: 0

Views: 1045

Answers (1)

talat
talat

Reputation: 70286

Although it is not the most efficient code on earth, you could use a nested ifelse:

df$final <- with(df, ifelse(is.na(flag), original, 
                            ifelse(flag, pmax(original, corrected), corrected)))
df
#  original corrected flag final
#1       20        25    0    25
#2       40        43    0    43
#3       20        22    0    22
#4        2         1    1     2
#5       20        25    1    25
#6       10         7    1    10
#7       12        NA   NA    12

You can make use of the fact that flag can be interpreted as a logical value.

Upvotes: 1

Related Questions