Count non-null variables and put into a new columb by grouping

Asked May 10 '18 at 09:43

Active May 10 '18 at 09:46

Viewed 28 times

Here is head of my dataset as follows:

ID    L.NO   Root1  Root2  Root3
1     2120   SX     NA     NA
2     2121   NC     NA     NA
3     2122   NA     NA     SX
4     2223   NA     CX     PX
5     2224   NA     SX     LX

I'd like to count non-null variables and put into a new column by grouping ID. How can I do that in R?

ID    L.NO   Root1  Root2  Root3 NewCol
1     2120   SX     NA     NA    2
2     2121   NC     NA     NA    2
3     2122   NA     NA     SX    2
4     2223   NA     CX     PX    1
5     2224   NA     SX     LX    1

edited May 10 '18 at 09:46

Arnab

4,216
2
28
50

asked May 10 '18 at 09:43

kimi

2

Why is the 2nd row, `NewCol` is `2` ? Are we grouping by `ID` column? In this example all IDs are unique, please provide a better example. – zx8754 May 10 '18 at 09:52
Try `rowSums(!is.na(df))` – Sotos May 10 '18 at 10:01
Because there is 2 NA which belongs to ID 1. – kimi May 10 '18 at 10:08
rowSums(!is.na(df)) works but how can I put the results to a new column grouping by ID? – kimi May 10 '18 at 10:12
`ds$NewCol <- rowSums(is.na(df)) ` – chinsoon12 May 10 '18 at 10:12
Yeah, thanks. I missed the point :) – kimi May 10 '18 at 10:21

Count non-null variables and put into a new columb by grouping

0 Answers0