Create multiple dataframes and add a column to them within a loop

Question

I'd like to create multiple data frames with a for loop, and add a column and remove others from each of the new data frames. The original dataset looks something like this:

Site  Count1  Count2  Count3  Count4  Count5  Count6  Count7  Count8  Count9  Habitat
  1     0       0       0       0       1       0        2       0       1     Forest
  2     1       2       3       0       0       2        0       1       0     Field
  3     2       0       0       1       1       1        0       2       0     Field

Essentially I want to create a new data frame for each count column, so here I would have nine new dataframes that look like this:

df.1:

Site TotCount Habitat
  1     0       Forest
  2     1       Field
  3     2       Field

df.2:

Site TotCount Habitat
  1     0       Forest
  2     2       Field
  3     0       Field

And so on.

This creates the frames I need:

for (z in 1:9){       
  assign(paste("df",z,sep="."),orig_data)
}

I can't figure out the second step - carry the appropriate count into each new data frame along with the habitat column (and others not shown above). I am hoping to keep this automated as it's possible there could be many more than 9 counts.

flodel · Accepted Answer · 2013-10-28 22:04:28Z

4

This will create a list of nine data.frames which is (trust me) a much better thing than creating nine variables into your global environment:

data.list <- lapply(1:9, function(i) {
  setNames(data[c("Site", paste0("Count", i), "Habitat")],
           c("Site", "TotCount", "Habitat"))
})

Instead of df.1, you can access the first data.frame by doing data.list[[1]]. Also, if you want to apply a function to all data.frames (i.e. a loop), have a look at lapply or sapply.

answered Oct 28, 2013 at 22:04

flodel

89.3k21 gold badges190 silver badges225 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

ricardo · Accepted Answer · 2013-10-28 22:16:37Z

This will put the df.x objects into the global environment.

# example data
df <- data.frame(Site = 1:3, 
                 Count1 = 0:2, 
                 Count2 = c(0,2,0), 
                 Count3 = c(0,3,0), 
                 Habitat = c("Forest", "Field", "Field") 
                 )

A function to do the work:

countSplit_2df <- function(DF){
    CountCols <- grep("Count", names(DF))
    HabCol <- grep("Habitat", names(DF))
    for (z in CountCols) {
        assign(paste("df", z, sep="."), 
               rbind(df[,c(1, z, HabCol)]),
               envir = .GlobalEnv)
    }
}

Now apply it:

countSplit_2df(df)

And ls() and you'll see

> ls()
[1] "countSplit_2df" "df"             "df.2"           "df.3"          
[5] "df.4"

Note this is not exactly as you asked for, as i've named them using the column number from the grep.

To complete the answer, someone needs to get the right way of selecting the text after "Count" in the grep -- you'd then use that to label the df.x objects. I couldn't get it to work.

Collectives™ on Stack Overflow

Create multiple dataframes and add a column to them within a loop

2 Answers 2

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related