Using for statement to compare plots in a dataframe in R

Question

I am trying to compare and plot a variable's distribution with that of the variable's log transformation.

What I am saying below, for every variable in mtcars if it is either a numeric or an integer return a histogram of it, then return a histogram with a log transformation, so that I can compare.

Any help would be appreciated.

for(i in ncol(mtcars)){
   par(mfcol = c(1,2))
   if (as.numeric | as.integer(mtcars[,i]) == T){
      return(hist(mtcars[,i]))}
   if (as.numeric | as.integer(mtcars[,i]) == T){
      return(hist(log(mtcars[,i])+1))}
}

Error in as.numeric | as.integer(mtcars[, i]) == T : 
operations are possible only for numeric, logical or complex types

Answer 1

This has nothing to do with hist() , it's the if statement that does not make much sense.

• You want to use is.numeric() and is.integer()
• Both require the argument is.numeric(mtcars[,i]) and is.integer(mtcars[,i]) .
• is.numeric() and is.integer() already return a boolean , so there is no need to check == T

Your code should read:

for(i in ncol(mtcars)){
  if (is.numeric(mtcars[,i]) | is.integer(mtcars[,i])){
    return(hist(mtcars[,i]))
    return(hist(log(mtcars[,i])+1)}
}

You should also know that it's almost always better to leverage the apply functions family instead of loops, eg:

apply(mtcars, 2, function(x) {hist(log(x)+1); hist(x)})

Answer 2

You should use function is.numeric and is.integer . Using as. makes no sense in an if statement.

This would be the correct approach:

for(i in ncol(mtcars)){
    par(mfcol = c(1,2))
    if (is.numeric(mtcars[,i] | is.integer(mtcars[,i])){
       return(hist(mtcars[,i]))
    }
    else {
       return(hist(log(mtcars[,i])+1))
    }
}