CODEWITHSUNDEEP
r
Arduin
Arduin

Reputation: 233

function to loop string column to numeric in R

I'm trying to convert few column of a data frame from char to numeric in R.

Exemple:

transform_function <- function(dataframe, vector){
  for (i in vector){
    dataframe[[i]] <- as.numeric(dataframe[[i]])
  }
  return(dataframe)
}

where my data frame has columns from a to z and I do this:

vector <- c("a", "c", "d", "h")

transform_function(dataframe, vector)

Getting the follow error:

Error in `[[<-.data.frame`(`*tmp*`, i, value = numeric(0)) : 
  replacement has 0 rows, data has 11068 
4.
stop(sprintf(ngettext(N, "replacement has %d row, data has %d", 
    "replacement has %d rows, data has %d"), N, nrows), domain = NA) 
3.
`[[<-.data.frame`(`*tmp*`, i, value = numeric(0)) 
2.
`[[<-`(`*tmp*`, i, value = numeric(0)) 
1.
transform_function(dataframe, vector)

Upvotes: 1

Views: 1244

Answers (3)

Rushabh Patel
Rushabh Patel

Reputation: 2764

Most efficient way:

 cols_to_convert <- colnames(dt)
 data.table::setDT(dt)[, (cols_to_convert) := lapply(.SD, as.numeric), .SDcols = cols_to_convert ]

Upvotes: 0

IceCreamToucan
IceCreamToucan

Reputation: 28695

You can use dplyr::mutate_at.

library(dplyr)
to_change <- c('a', 'c')

df <- data.frame(a = as.character(1:10), b = as.character(11:20), c = as.character(21:30))
df %>% mutate_at(to_change, as.numeric)

Upvotes: 1

Rilcon42
Rilcon42

Reputation: 9763

Try using apply to iterate over the columns rather than writing your own function. Example solution:

df <- data.frame(c("a", "c", "d", "h"),c(1,2,3,4))

df<-apply(df,2,as.numeric)
df

Upvotes: 1

Related Questions