Is there a way to initialize or modify a list that allows to reference previous values?-CodePudding

I want to store non-rectangular data in a structure, e.g. in a list.

Referencing previous data works using a tibble, e.g. like this:

dat <- tibble(a = 2, 
              b = list(c(a   1, a   2)),
              c = list(tibble(a = a, b = 3)))

Note that because tibble does not allow tibble in columns (otherwise the data would not be rectangular), we have to wrap the tibble in a list and dat$c does not return the tibble, for that we need dat$c[[1]].

The latter problem can be solved with a list.

dat2 <- list(a = 2, 
             b = c(3, 4), 
             c = tibble(a = 2, b = 3))

However, now I can't construct the data element with reference to previous entries:

dat2 <- list(a = 2, 
             b = c(a   1, a   2), 
             c = tibble(a = a, b = 3))
> Error: object 'a' not found

My question: Is there a way to construct a non-rectangular data structure that does support construction by reference to previous entries?

CodePudding user response：

You can use lst() instead:

lst(a = 2, 
    b = c(a   1, a   2), 
    c = tibble(a = a, b = 3))

$a
[1] 2

$b
[1] 3 4

$c
# A tibble: 1 × 2
      a     b
  <dbl> <dbl>
1     2     3

CodePudding user response：

You can use a = (a <- 2) for the reference

dat <- tibble(
    a = (a<-2),
    b = list(c(a   1, a   2)),
    c = list(tibble(a = a, b = 3))
)

and you will see

> str(dat)
tibble [1 x 3] (S3: tbl_df/tbl/data.frame)
 $ a: num 2
 $ b:List of 1
  ..$ : num [1:2] 3 4
 $ c:List of 1
  ..$ : tibble [1 x 2] (S3: tbl_df/tbl/data.frame)
  .. ..$ a: num 2
  .. ..$ b: num 3