Remove a row from all dataframes in a list if NA value in one of the rows

user_10192633 0
Viewed 24

I have a list of data.frames of equal size. There exist missing data in different rows and columns of each data.frame. I would like to remove the row of each data frame for which one of data.frames have a row that contains a NaN. The current lapply and na.omit code I have removes each row corresponding to the specific data.frame which makes sense as it goes through each data.frame in the list before moving on to the next one. However, I would like to make it so that if an NaN exists in one row of a data.frame that row gets removed from all other data.frames

Some example code:

#Make list
ls <- list(x1=data.frame(a=c(1,2,3,4),b=c(2,3,4,5),c=c(3,4,NaN,6)),
           x2=data.frame(a=c(1,NaN,3,4),b=c(2,3,4,5),c=c(3,4,5,6)))
#Desired output
lscalc <- list(x1=data.frame(a=c(1,4),b=c(2,5),c=c(3,6)),
               x2=data.frame(a=c(1,4),b=c(2,5),c=c(3,6)))
dataframe nan r list
2 Answers

Assuming all the datasets have the same number of rows, get the row index from all the datasets first and then loop over the list and remove those rows

un1 <- unique(unlist(lapply(ls, function(x) which(is.na(x), arr.ind = TRUE)[,1])))
lapply(ls, function(x) x[!seq_len(nrow(x)) %in% un1, ])
$x1
  a b c
1 1 2 3
4 4 5 6

$x2
  a b c
1 1 2 3
4 4 5 6
user_37322710

Here's one using complete.cases(), though otherwise along the same lines as @akrun's.

#Make list
l <- list(x1=data.frame(a=c(1,2,3,4),b=c(2,3,4,5),c=c(3,4,NaN,6)),
           x2=data.frame(a=c(1,NaN,3,4),b=c(2,3,4,5),c=c(3,4,5,6)))
#Desired output
lcalc <- list(x1=data.frame(a=c(1,4),b=c(2,5),c=c(3,6)),
               x2=data.frame(a=c(1,4),b=c(2,5),c=c(3,6)))

inds <- lapply(l, \(x)which(!complete.cases(x)))
inds <- unique(do.call(c, inds))
lcalc2 <- lapply(l, \(x)x[-inds, ])
lcalc2
#> $x1
#>   a b c
#> 1 1 2 3
#> 4 4 5 6
#> 
#> $x2
#>   a b c
#> 1 1 2 3
#> 4 4 5 6

Created on 2022-05-24 by the reprex package (v2.0.1)

user_82064340
Related