I have a list of vectors and I wish to check them and add the ones that are not empty into a data frame. My main issue when doing this is keeping the vector names so I can then add them. My data is as follows;

att1 = 5
att2 = 4
att3 = character()
att4 = 8

scoresDf = data.frame(att1,att2,att3,att4)

This cannot be made into a dataframe as att3 in this case is empty. Therefore I need to find out which of my columns is empty and only add the ones to the df what are length()>0. When I test which vectors are length>0 my issue is i cant keep the vector names to insert into my data frame.

What I am trying to do is something like;

 goodData = #my vectors that are not empty
 scoresDf = data.frame(goodData)

I have a list of vectors and I wish to check them and add the ones that are not empty into a data frame. My main issue when doing this is keeping the vector names so I can then add them. My data is as follows;

att1 = 5
att2 = 4
att3 = character()
att4 = 8

scoresDf = data.frame(att1,att2,att3,att4)

This cannot be made into a dataframe as att3 in this case is empty. Therefore I need to find out which of my columns is empty and only add the ones to the df what are length()>0. When I test which vectors are length>0 my issue is i cant keep the vector names to insert into my data frame.

What I am trying to do is something like;

 goodData = #my vectors that are not empty
 scoresDf = data.frame(goodData)

• r

• We can only collect in a data.frame what is of same length (exception list columns). You can do lengths(mget(ls(pattern='^att\\d+$'))) to check lengths. empty is of length zero here. I.e. L = mget(ls(pattern='^att\\d+$')); X = list2DF(L[lengths(L)==1]); str(X) – Friede Commented 21 hours ago
• @Friede Yeh thats what I'm trying to say, how do I identify which are empty and add the non-empties into my Df – Joe Commented 20 hours ago
• Are all non-empty vectors of same length? – Friede Commented 20 hours ago
• @Friede Yep either 1 or 0 – Joe Commented 20 hours ago

3 Answers 3

You might want to make your toy data more real.

1) Exactly one element

att1 = 5
att2 = 4
att3 = character()
att4 = 8

L = mget(ls(pattern='^att\\d+$'))
X = list2DF(L[lengths(L)==1])

> str(X)
'data.frame':   1 obs. of  3 variables:
 $ att1: num 5
 $ att2: num 4
 $ att4: num 8
> X
  att1 att2 att4
1    5    4    8

2) At least one element

att1 = 5
att2 = 4
att3 = character()
att4 = 8
att5 = 1:2

L = mget(ls(pattern='^att\\d+$')) |> 
  Filter(f=length)  
lapply(L, `length<-`, max(lengths(L))) |> 
  list2DF()

  att1 att2 att4 att5
1    5    4    8    1
2   NA   NA   NA    2

Note

If length cannot be different from 0 or 1, why not returning a (named) list or numericvector instead of a data.frame (list of vectors)?

att1 = 5
att2 = 4
att3 = character()
att4 = 8
#att5 = 1:2

> mget(ls(pattern='^att\\d+$')) |> Filter(f=length)
$att1
[1] 5

$att2
[1] 4

$att4
[1] 8
> # or 
> mget(ls(pattern='^att\\d+$')) |> Filter(f=length) |> unlist()
att1 att2 att4 
   5    4    8 

att1 = 5
att2 = 4
att3 = character()
att4 = 8

scoresDf = list(att1=att1,
                att2=att2,
                att3=att3,
                att4=att4)
hasData <- sapply(scoresDf, function(x) {length(x) != 0} )

data.frame(scoresDf[hasData])

returns:

> data.frame(scoresDf[hasData])
  att1 att2 att4
     5    4    8

Another implementation is using subset + lengths to filter out the empty entries, e.g.,

mget(ls(pattern = "^att\\d"))  %>% 
  subset(!!lengths(.)) %>% 
  list2DF()