R: split a data.frame but keep classes

user_8061552 0
Viewed 108

I would like to split a dataframe into a list of dataframes and keep the classes of the variables.

# create sample data
df <- data.frame(
  id=c("1","2"),
  site_name = c("Zero Hedge", "Free Software Foundation"),
  site_url = c("https://www.zerohedge.com", "https://www.fsf.org")
)

# specify class for site_url 
class(df$site_url) <- "formula"

# split
dataframes <- split(df, df$id)

Now I wonder, why the splitted data changed the class:

class(dataframes[[1]]$site_url)
[1] "character"

My questions:

  1. Why does that happen?
  2. How can I split a dataframe into a list of dataframes and keep the classes of the variables?

Thank you for your help.

Additional info:

I came across this problem when I wanted to automatically write hyperlinks to excel files with R and openxlsxaccording to this very helpful post: Openxlsx hyperlink output display in Excel

dataframe r class split
1 Answers

We can set the attributes

dataframes2 <- lapply(dataframes, function(x) {
      attributes(x$site_name) <- attributes(df$site_name)
     x})

The issue is not related to split or methods of it. In this case, it is split.data.frame. If we look at the source code, it is splitting based on the sequence of rows based on the grouping 'f' and then doing the extraction ([)

split.data.frame
function (x, f, drop = FALSE, ...) 
lapply(split(x = seq_len(nrow(x)), f = f, drop = drop, ...), 
    function(ind) x[ind, , drop = FALSE])

But, the split.data.table keeps the class

split(as.data.table(df), df$id) %>% str
#List of 2
# $ 1:Classes ‘data.table’ and 'data.frame':    1 obs. of  3 variables:
# ..$ id       : chr "1"
# ..$ site_name: chr "Zero Hedge"
# ..$ site_url : 'formula' chr "https://www.zerohedge.com"
 # ..- attr(*, ".internal.selfref")=<externalptr> 
# $ 2:Classes ‘data.table’ and 'data.frame':    1 obs. of  3 variables:
#  ..$ id       : chr "2"
#  ..$ site_name: chr "Free Software Foundation"
#  ..$ site_url : 'formula' chr "https://www.fsf.org"

-checking the structure of the original data with the extracted rows data

str(df)
'data.frame':   2 obs. of  3 variables:
 $ id       : chr  "1" "2"
 $ site_name: chr  "Zero Hedge" "Free Software Foundation"
 $ site_url : 'formula' chr  "https://www.zerohedge.com" "https://www.fsf.org" str(df[1,]) # with one row selected
'data.frame':   1 obs. of  3 variables:
 $ id       : chr "1"
 $ site_name: chr "Zero Hedge"
 $ site_url : chr "https://www.zerohedge.com" # lost attribute
str(df[1:2,]) # with more than one row
'data.frame':   2 obs. of  3 variables:
 $ id       : chr  "1" "2"
 $ site_name: chr  "Zero Hedge" "Free Software Foundation"
 $ site_url : chr  "https://www.zerohedge.com" "https://www.fsf.org" # lost attribute
user_37322710
Related