R: converting fractions into decimals in a data frame

问题

I am trying to convert a data frame of numbers stored as characters in a fraction form to be stored as numbers in decimal form. (There are also some integers, also stored as char.) I want to keep the current structure of the data frame, i.e. I do not want a list as a result.

Example data frame (note: the real data frame has all elements as character, here it is a factor but I couldn't figure out how to replicate a data frame with characters):

    a <- c("1","1/2","2")
    b <- c("5/2","3","7/2")
    c <- c("4","9/2","5")
    df <- data.frame(a,b,c)

I tried df[] <- apply(df,1, function(x) eval(parse(text=x))). This calculates the numbers correctly, but only for the last column, populating the data frame with that.

Result:

   a  b    c
1  4  4.5  5
2  4  4.5  5
3  4  4.5  5

I also tried df[] <- lapply(df, function(x) eval(parse(text=x))), which had the following result (and I have no idea why):

Desired result:

   a   b    c
1  1   2.5  4
2  0.5 3    4.5
3  2   3.5  5

Thanks a lot!

回答1:

You are probably looking for:

df[] <- apply(df, c(1, 2), function(x) eval(parse(text = x)))
df
    a   b   c
1 1.0 2.5 4.0
2 0.5 3.0 4.5
3 2.0 3.5 5.0

eval(parse(text = x))

evaluates one expression at a time so, you need to run cell by cell.

EDIT: if some data frame elements can not be evaluated you can account for that by adding an ifelse statement inside the function:

df[] <- apply(df, c(1, 2), function(x) if(x %in% skip){NA} else {eval(parse(text = x))})

Where skip is a vector of element that should not be evaluated.

回答2:

Firstly, you should prevent your characters from turning into factors in data.frame()

df <- data.frame(a, b, c, stringsAsFactors = F)

Then you can wrap a simple sapply/lapply inside your lapply to achieve what you want.
```
sapply(X = df, FUN = function(v) {
                              sapply(X = v,
                                     FUN = function(w) eval(parse(text=w)))
                             }
  )
```
Side Notes
If you feed eval an improper expression such as expression(1, 1/2, 2), that evaluates to last value. This explains the 4 4.5 5 output. A proper expression(c(1, 1/2, 2)) evaluates to the expected answer.
The code lapply(df, function(x) eval(parse(text=x))) returns a 3 3 2 because sapply(data.frame(a,b,c), as.numeric) returns:
```
     a b c
[1,] 1 2 1
[2,] 2 1 3
[3,] 3 3 2
```
These numbers correspond to the levels() of the factors, through which you were storing your fractions.

来源：https://stackoverflow.com/questions/48955593/r-converting-fractions-into-decimals-in-a-data-frame

标签

dataframe

apply

lapply

fractions