Use purrr::accumulate with condition

Question

Below is a reproducible example with a test dataframe:

dat<- structure(list(A = c(1.3, 1.5, 1.6, 1.2, 1.1, 1.2), 
                     B = c(0.25, 0.21, 0.21, 0.15, 0.26, 0.17)), 
                class = c("tbl_df", "tbl", "data.frame"), 
                row.names = c(NA, -6L))

I want to do add a column with an initial value, say 1000, and then use accumulate but depending on a condition on column A.

Basically, if value in column A is superior or equal to 1.2; Column C = 1000-(Column B * 1000). But this is for the first row only. For the rest, instead of 1000 (initial value), it should be previous value.

My desired output would be something like this :

A	B	C
1.3	0.25	750
1.5	0.21	592.5
1.6	0.21	468.075
1.2	0.15	397.864
1.1	0.26	397.864
1.2	0.17	330.227

(and so the first row should be 1000 if column A is inferior to 1.2).

I've tried to use accumulate but I can't make the conditional part work:

dat<-dat %>%
  mutate(C = accumulate(tail(B,-1), .f = ~ .x - (.x * .y), .init = 1000))

Answer 1

If you would like to use accumulate

> dat %>%
+     mutate(C = accumulate(B * (A >= 1.2), ~ .x * (1 - .y), .init = 1000)[-1])
# A tibble: 6 × 3
      A     B     C
  <dbl> <dbl> <dbl>
1   1.3  0.25  750
2   1.5  0.21  592.
3   1.6  0.21  468.
4   1.2  0.15  398.
5   1.1  0.26  398.
6   1.2  0.17  330.

---

With base R, you can do the same if you use Reduce or cumprod

transform(
    dat,
    C = Reduce(\(x, y) x * (1 - y),
        B * (A >= 1.2),
        init = 1000,
        accumulate = TRUE
    )[-1]
)

or

transform(
    dat,
    C = cumprod(1 - B * (A >= 1.2)) * 1000
)

which gives

    A    B        C
1 1.3 0.25 750.0000
2 1.5 0.21 592.5000
3 1.6 0.21 468.0750
4 1.2 0.15 397.8638
5 1.1 0.26 397.8638
6 1.2 0.17 330.2269

data

> dput(dat)
structure(list(A = c(1.3, 1.5, 1.6, 1.2, 1.1, 1.2), B = c(0.25, 
0.21, 0.21, 0.15, 0.26, 0.17)), row.names = c(NA, -6L), class = c("tbl_df",
"tbl", "data.frame"))