CODEWITHSUNDEEP
r
user1471980
user1471980

Reputation: 10646

how do you add new columns to an empty data frame in R

I declare an empty data frame as this:

df <- data.frame()

then I go though processing some files and as process, I need to build my df data frame. I need to keep adding columns to it:

For example, I process some file and build a data frame called new_df, I now need to add this new_df to my df:

I've tried this:

latest_df <- cbind(latest_df, new_df)

I get this error:

Error in data.frame(..., check.names = FALSE) : arguments imply differing number of rows: 0, 1

Upvotes: 8

Views: 13269

Answers (2)

Eva
Eva

Reputation: 1

If you simply want an empty data frame of the proper size before you enter the loop, and assuming "df" and "new_df" have the same number of rows x, try

df <- data.frame(matrix(nrow=x))

for (i in 1:n){
temp[i] <- % some vector of length x
}

Upvotes: 0

ChristyCasey
ChristyCasey

Reputation: 446

Just put data into the index after the last column

new_df = data.frame()


new_df[,ncol(new_df)+1] = NA

So if you knew you had 3 columns then:

new_df[,4] = c('a','b','c')

Example:

new_df = data.frame('a'=NA)
for(i in 1:10){
  new_df[,ncol(new_df)+1] = NA
}
new_df

EDIT:

ProcessExample <- function(){
  return(c(5)) #just returns 5 as fake data everytime
}

new_df = data.frame(matrix(nrow=1))
for(i in 1:10){
  new_df[,ncol(new_df)+1] = ProcessExample()
}
latest_df <- new_df[,-1]

Or just add rows and transpose the data set

new_df = data.frame()
for(i in 1:10){
  new_df[i,1] = ProcessExample()
}
latest_df <- t(new_df)

Upvotes: 5

Related Questions