Replace column with another column

Question

I would like to replace the value of column Name.x with value from Name.y with condition if it is not NA (empty rows)

Name.x  Name.y
US      NA
Germany NA
Germany France
Canada  NA
Italy     Morocco
Austria Belgium

Result:

Name.x
US
Germany
France
Canada
Morocco
Belgium

do a `coalesce` i.e. `library(dplyr);d1 %>% mutate(Name.x = coalesce(Name.x, Name.y)) %>% select(Name.x)` — akrun, Feb 15 '18 at 10:44
If it is only 2 columns then try: `df1$Name.x <- ifelse(!is.na(df1$Name.y), df1$Name.y, df1$Name.x)` — zx8754, Feb 15 '18 at 10:53
it is a big dataset... this is just an example to make it simplier — user5879741, Feb 15 '18 at 10:54
Data can be big, but are you comparing only 2 columns or more? If more, then use coalesce as suggested by akrun. — zx8754, Feb 15 '18 at 10:56
i only want name.x column in this dataframe1 to be basically overwritten by dataframe2 only with condition that it is not empty. — user5879741, Feb 15 '18 at 11:05
Possible duplicate of [How to implement coalesce efficiently in R](https://stackoverflow.com/questions/19253820/how-to-implement-coalesce-efficiently-in-r) — Aramis7d, Feb 15 '18 at 12:10
@akrun I think it should be `coalesce(Name.y, Name.x)`, going by the example provided. — Aramis7d, Feb 15 '18 at 12:12
@Aramis7d Could be, I think there are some edits in the example — akrun, Feb 15 '18 at 13:47

score 1 · Answer 1 · edited Feb 15 '18 at 12:25

1

Example data:

a <- data.frame("Name.x" = c("US", "Germany","Germany", "Canada", "Italy", "Austria"),  "Name.y" = c(NA, NA, "France", NA, "Morocco", "Belgium"))

Solution:

a$Name.x <- ifelse(is.na(a$Name.y), as.character(a$Name.x), as.character(a$Name.y))

edited Feb 15 '18 at 12:25

duckmayr

16,303
3
35
53

answered Feb 15 '18 at 11:31

R.Y

21
1

1

Welcome to `stackoverflow`. Please always try to explain what you're doing in the code. It's not really helpful to just submit unformatted code without any explanations. – Aramis7d Feb 15 '18 at 12:09
1

This is a good answer. The downvote is a bit harsh here especially considering that the code is self explanatory. – Sotos Feb 15 '18 at 13:35

Terru_theTerror · Accepted Answer · 2018-02-15T13:29:29.610

0

Try something like this:

Your data.frame

 db1<-data.frame(Name.x=c("US","Germany"),
                 Name.y=c(NA,"France"))
   db1
   Name.x Name.y
      US   <NA>
 Germany France

Columns names to analyze/substitute

coldb1_NA<-"Name.y"
coldb_toSub<-"Name.x"

Substitution

db2<-db1
db2[,coldb_toSub]<-as.character(db1[,coldb_toSub])
db2[!is.na(db1[,coldb1_NA]),coldb_toSub]<-as.character(db1[!is.na(db1[,coldb1_NA]),coldb1_NA])

Your output

db2
  Name.x Name.y
1     US   <NA>
2 France France

edited Feb 15 '18 at 13:29

answered Feb 15 '18 at 10:53

Terru_theTerror

4,918
2
20
39

it is a big datasets with different other names ... is there a way to wrtie something like this ... total_split[, "Name.x"] <- total_split[, "Name.y"=!''] ? – user5879741 Feb 15 '18 at 11:00
This code is not relied to the names of the countries present inside che two columns (Name.x, Name.y). You want that this code works also on other columns? – Terru_theTerror Feb 15 '18 at 11:05
Added a general solution with parametrization of columns names – Terru_theTerror Feb 15 '18 at 11:25
what the db1 and db2 stand for ? – user5879741 Feb 15 '18 at 12:52
i replace db1 and db2 by the name of my dataframe and basically the result i got , is that Name.y overworte Name.x including empty rows as well ... – user5879741 Feb 15 '18 at 13:05
Name.x in db2 replaces Name.x in db1 where Name.y in db1 is NA. Isn't this what you want? – Terru_theTerror Feb 15 '18 at 13:13
yes exactly... but I dont want NA in Name.y to overwrite Name.x ... this mean Name.y overwrite Name.x ONLY when Name.y have a name .. – user5879741 Feb 15 '18 at 13:16
Ok, look at the answer now. – Terru_theTerror Feb 15 '18 at 13:29

Replace column with another column

2 Answers2