docs: 📝 update filter example in guide (#274)

# Description

~~`foed_dato` is now a `date` in the simulated data, so the mutate to
`strptime()` started failing locally.~~

Edit: Regenerating `koen` and using that in the filter example, so we
have full control of the type regardless of potential future changes
upstream.

Needs a quick review.

## Checklist

- [X] Ran `just run-all`

Author: signekb

vignettes/fastreg.qmd

@@ -55,7 +55,12 @@ bef_list <- simulate_register(
   "bef",
   c("", "1999", "1999_1", "2020"),
   n = 1000
-)
+) |>
+  # Randomly (re)generate koen, so we don't rely on any potential future
+  # changes to the simulated data from osdc.
+  purrr::map(\(x) {
+    x |> dplyr::mutate("koen" = sample(c(1, 2), 1000, replace = TRUE))
+  })
 
 lmdb_list <- simulate_register(
   "lmdb",
@@ -218,14 +223,11 @@ file
 ```
 
 The resulting DuckDB table can be filtered and transformed with `dplyr`.
-For example, `foed_dato` is stored as a character in the simulated data
-(`"YYYYMMDD"`), so you can change it to a date (using `strptime()`) and
-then filter:
+For example, you can filter the data:
 
 ```{r filter-file}
 file |>
-  dplyr::mutate(foed_dato = strptime(foed_dato, "%Y%m%d")) |>
-  dplyr::filter(koen == 2 & foed_dato > "2000-01-01") |>
+  dplyr::filter(koen == 2) |>
   dplyr::compute()
 ```