Maintaining names while flattening lists in R

Question

Below is some dummy data and code. I think I'm trying to do the exact opposite to this question.

After manipulating for a statistical test, I'm left with a very ugly looking list. When I unpack this list as per this method with unlist, the variable/country coding stored in names is totally lost.

I would like to have a data frame with countries as row names and categories as column names.

Any pointers would be much appreciated!

set.seed(1)

CVtest <- list()

for (i in c("AT","BE","DE")) { #For all countries

  # Test all categories together

  CVtest[[i]][["EveryCat"]] <-  sample(seq(0, 1, by=0.01), 1)

    for (j in c("All", "Sub", "Key", "Sel")) { # Test each category

      CVtest[[i]][[j]] <- sample(seq(0, 1, by=0.01), 1)
    }
}

Current output:

> class(CVtest)
[1] "list"
> CVtest

$`AT` EveryCat      All      Sub      Key      Sel 
    0.26     0.37     0.57     0.91     0.20 

$BE EveryCat      All      Sub      Key      Sel 
    0.90     0.95     0.66     0.63     0.06 

$DE EveryCat      All      Sub      Key      Sel 
    0.20     0.17     0.69     0.38     0.77

Desired output

> class(CVtest)
[1] "data.frame"
> CVtest
       EveryCat  All  Sub  Key  Sel
AT     0.26 0.37 0.57 0.91 0.20
BE     0.90 0.95 0.66 0.63 0.06
DE     0.20 0.17 0.69 0.38 0.77

Can't you do a minimal example ? You provide some code but it is lengthy and it requires to install some packages. — Stéphane Laurent, Aug 28 '18 at 12:41
You don't provide `df` in your post. And the index `i` of the loop is not used. Also, could you show the desired output? — Stéphane Laurent, Aug 30 '18 at 10:31
sorry again, `i` was indeed meant to replace `df`... forgot about that bit. Now fixed including desired output — Fons MA, Sep 01 '18 at 05:24

score 0 · Accepted Answer · answered Aug 30 '18 at 10:44

0

Your question is still not clear (see my latest comment). I can only guess what you want. Could you try:

x <- purrr::transpose(CVtest)
y <- rlist::list.stack(x)
rownames(y) <- names(x)
y

Say me whether y is the desired output. Otherwise I will delete this answer and please clarify your question.

answered Aug 30 '18 at 10:44

Stéphane Laurent

75,186
15
119
225

I fixed the code and other parts of the question; I believe it's quite clear right now. Thanks for your patience. Your solution does the job, I wanted to have countries as rows, so I just added t(y). Thank you very much!! – Fons MA Sep 01 '18 at 05:27

Maintaining names while flattening lists in R

1 Answers1