Mutate IF ELSE MULTIPLE COLUMNS-CodePudding

I have a file (df) with two columns (LAB and OID)... I would like to create a new variable on df file: column3(OID_new), using the information of Columns 1(LAB) and 2 (OID):

LAB      OID OID_new
12       NA    12
13       NA    13
14       NA    14
plate_1  18    18
plate_2  24    24
Plate_3  23    23

for this, i tried:

df %>%
 mutate(OID_new= ifelse(OID == NA,df$LAB,
                    ifelse(OID ^= NA,"df$OID")))-> df1

But don't work

Thanks in advance

CodePudding user response：

You can achieve that by using ifelse(), as you tried. A few things to mention:

ifelse() requires three statements:
1. a test (here: check if OID is NA: is.na(OID)
2. an output if your test result is yes: return LAB
3. an output if your test result in no: return OID
if you pipe %>% your dataframe, there's no need to access your dataframe by using $
the base ifelse() is better here, since its dplyr version if_else() is type stable and would throw an error. This might be more convenient, but there might be unexpected results after implicit coercing, so be careful.

Code

df %>% mutate(OID_new = ifelse(is.na(OID), LAB, OID)) -> df1

Output

> df1
      LAB OID OID_new
1      12  NA      12
2      13  NA      13
3      14  NA      14
4 plate_1  18      18
5 plate_2  24      24
6 plate_3  23      23

Data

df <- structure(list(LAB = c("12", "13", "14", "plate_1", "plate_2", "plate_3"), 
                     OID = c(NA, NA, NA, 18, 24, 23)), 
                class = "data.frame", 
                row.names = c(NA, -6L))