CODEWITHSUNDEEP
r
muyueh
muyueh

Reputation: 1038

Is it possible to reference other var in `transform`?

Is it possible to reference other var in transform?

For instance: 

Data = transform(Data, 
              m = x,
              d = m,
              c = m * 2
)

Upvotes: 1

Views: 55

Answers (2)

G. Grothendieck
G. Grothendieck

Reputation: 269556

Here are some solutions. For reproducibility we will assume this definition for Data:

Data <- data.frame(x = 1)

1) iterated transform Try an iterated transform:

transform(transform(Data, m = x), d = m, c = m*2)

2) within or use within:

within(Data, { m <- x; d <- m; c <- m*2 })

3) my.transform There is also an alternate transform defined here that supports such behavior.

my.transform(Data, m = x, d = m, c = m*2)

Note that this is the only answer here that supports setting the variables out of order, e.g.

my.transform(Data, d = m, c = m*2, m = x)

4) dplyr::mutate mutate in the dplyr package also supports this:

library(dplyr)
mutate(Data, m = x, d = m, c = m*2)

Upvotes: 5

Gavin Simpson
Gavin Simpson

Reputation: 174813

No, if you mean a variable you create in the current transform() call:

df <- data.frame(x = 1:10)
df2 <- transform(df, m = x, d = m, c = m * 2)

> df <- data.frame(x = 1:10)
> df2 <- transform(df, m = x, d = m, c = m * 2)
Error in eval(expr, envir, enclos) : object 'm' not found

The obvious solution is to do two transform() calls:

df2 <- transform(df, m = x)
df2 <- transform(df2, d = m, c = m * 2)
head(df2)

> head(df2)
  x m d  c
1 1 1 1  2
2 2 2 2  4
3 3 3 3  6
4 4 4 4  8
5 5 5 5 10
6 6 6 6 12

or nest them

df3 <- transform(transform(df, m = x), d = m, c = m * 2)

> head(df3)
  x m d  c
1 1 1 1  2
2 2 2 2  4
3 3 3 3  6
4 4 4 4  8
5 5 5 5 10
6 6 6 6 12

The alternative is within(), which can do what you want but is slightly less user-friendly:

df4 <- within(df, {
    m <- x
    d <- m
    c <- m * 2
  })

> head(df4)
  x  c d m
1 1  2 1 1
2 2  4 2 2
3 3  6 3 3
4 4  8 4 4
5 5 10 5 5
6 6 12 6 6

One idiosyncrasy of this is that you don't have any control of the ordering of the new variables.

Note that within() works because the entire expression (the thing in { ... }) is evaluated within an environment containing the variables in df. Hence m is available in that environment as it is created in the first line of the expression.

Upvotes: 1

Related Questions