Calculate gaps using model frames

[davidskalinder]

Should fix #6. Hard to test as always, but here's a minimal-yet-still-clunky one:

library(OaxacaBlinder)

clna <- 
  chicago_long |>
  dplyr::mutate(educ2 = dplyr::na_if(education, "LTHS"))

twofold <-
  OaxacaBlinderDecomp(
    formula = real_wage ~ educ2 | female, 
    data = clna, 
    type = "twofold", 
    baseline_invariant = TRUE,
    n_bootstraps = 30
  )

clna |>
  dplyr::filter(!is.na(educ2)) |>
  dplyr::group_by(female) |>
  dplyr::summarize(mean = mean(real_wage, na.rm = TRUE)) |>
  dplyr::mutate(diff = dplyr::lag(mean) - mean) |>
  purrr::pluck("diff", 2) |>
  identical(twofold$gaps$gap)
#> [1] TRUE

^{Created on 2024-03-15 with reprex v2.1.0}