在 R 中使用 Sql 将 NA 替换为空白

Question

I am trying to replace NA with Blank using Sql in R. I used below code but I still get "NA" instead of blank!

DF_TEST_PN <- sqldf("Select *, Case when Proj_Loess ='NA' then ' ' else Proj_Loess end as New_Proj_Loess From Test")

Answer 1

You can use ifnull in sqldf / sqLite :

test <- tibble(Proj_Loess=c(NA,"test"))
DF_TEST_PN <- sqldf("Select *, ifnull(Proj_Loess,' ') as New_Proj_Loess From test")

  Proj_Loess New_Proj_Loess
1       <NA>               
2       test           test

Answer 2

Suppose we have DF as shown below. The NAs in R are represented by null in SQL so use coalesce as shown below. coalesce looks at each argument in turn and returns the first one that is not null, ie not NA. coalesce is available in all backends that sqldf supports (sqlite, h2, mysql, postgresql).

library(sqldf)

DF <- data.frame(A = c("A", NA, "B", NA, "C"))
sqldf("select *, coalesce(A, '') as V1 from DF")

giving:

     A V1
1    A  A
2 <NA>   
3    B  B
4 <NA>   
5    C  C

If you want to use case when as in the question then use is null like this:

sqldf("select *, case when A is null then '' else A end as V1 from DF")