[R] converting blank cells to NAs

[arun]

You can use na.strings="" in read.table() or read.csv()


library(stringr)

vec1<-unlist(str_split(readLines(textConnection("3,7,11,,12,14,15,,17,18,19")),","))
 vec1[vec1==""]<- NA
 vec1
# [1] "3"  "7"  "11" NA   "12" "14" "15" NA   "17" "18" "19"

#or
scan(text="3,7,11,,12,14,15,,17,18,19",sep=",")
#Read 11 items
 #[1]  3  7 11 NA 12 14 15 NA 17 18 19
set.seed(25)
 arr1<- array(sample(letters,40,replace=TRUE),dim=c(5,4,2))
arr1[5,3,2]<-""
arr1[arr1==""]<- NA
arr1[,,2]
#     [,1] [,2] [,3] [,4]
#[1,] "b"  "h"  "v"  "k" 
#[2,] "l"  "b"  "j"  "y" 
#[3,] "c"  "p"  "g"  "j" 
#[4,] "e"  "n"  "c"  "e" 
#[5,] "h"  "g"  NA   "d" 

 arr2<- array(1:40,dim=c(5,4,2))
 arr2[5,3,2]<- ""
arr2[arr2==""]<- NA
arr2<- apply(arr2,c(2,3),as.numeric)
arr2[,,2]
#     [,1] [,2] [,3] [,4]
#[1,]   21   26   31   36
#[2,]   22   27   32   37
#[3,]   23   28   33   38
#[4,]   24   29   34   39
#[5,]   25   30   NA   40


A.K.



>Is there a way to convert blank cells in a vector or array into "NA"s? 
 Alternatively, is there a test for blank cells the way one can use 
is.na to test for >NAs? 
>
>Thanks in advance.