Provide better error/recovery when test lags don't exist ("zero-length inputs [...]")

[brookslogan]

[Please see later comments for actual examples.]

[brookslogan]

The difference between this and #332 is that this is dealing with a lagset being bad for test data, while #332 is about a shiftset being bad for training data.

[dajmcdon]

Can you make a simple example without the slide?

[brookslogan]

Here's an example. Like with #332, I'm hoping for better error messages that spell out that I'm missing test-time predictors.

library(tibble)
library(dplyr)
library(workflows)
library(epiprocess)
library(epipredict)
# [message spam on the above removed]

edf <- tibble(
  geo_value = "ct",
  time_value = seq(as.Date("2020-10-01"), as.Date("2023-05-31"), by = "day"),
  ) %>%
  mutate(value = seq_len(nrow(.)) + rnorm(nrow(.))) %>%
  # Oct to May (flu season, ish) only:
  filter(!between(as.POSIXlt(time_value)$mon + 1L, 6L, 9L)) %>%
  # and actually, pretend we're around mid-October 2022:
  filter(time_value <= as.Date("2022-10-12")) %>%
  as_epi_df()

edf %>% arx_forecaster(
  "value",
  args_list = arx_args_list(lags = c(0L, 7L, 14L)) # (the current default)
)
#> Error in `as_list_of()`:
#> ! Can't find common type for elements of `x`.

edf %>% arx_forecaster(
  "value",
  trainer = quantile_reg(),
  args_list = arx_args_list(lags = c(0L, 7L, 14L)) # (the current default)
)
#> Error in (function (values = double(), quantile_levels = double()) : length(values) == length(quantile_levels) is not TRUE

^{Created on 2024-07-25 with reprex v2.1.1}

[#362 has another example.]