For any model that produces forecasts for multiple outcomes, such as multiple
aheads, the resulting prediction is a list of forecasts inside a column of
the prediction tibble, which is not an ideal format. This layer "lengthens"
the result, moving each outcome to a separate row, in the same manner as
tidyr::unnest() would. At the moment, the only such engine is
smooth_quantile_reg().
Usage
layer_unnest(frosting, ..., id = rand_id("unnest"))Arguments
- frosting
a
frostingpostprocessor- ...
<
tidy-select> One or more unquoted expressions separated by commas. Variable names can be used as if they were positions in the data frame, so expressions likex:ycan be used to select a range of variables.- id
a random id string
Examples
jhu <- covid_case_death_rates %>%
filter(time_value > "2021-11-01", geo_value %in% c("ak", "ca", "ny"))
aheads <- 1:7
r <- epi_recipe(jhu) %>%
step_epi_lag(death_rate, lag = c(0, 7, 14)) %>%
step_epi_ahead(death_rate, ahead = aheads) %>%
step_epi_naomit()
wf <- epi_workflow(
r,
smooth_quantile_reg(
quantile_levels = c(.05, .1, .25, .5, .75, .9, .95),
outcome_locations = aheads
)
) %>%
fit(jhu)
f <- frosting() %>%
layer_predict() %>%
layer_naomit() %>%
layer_unnest(.pred)
wf1 <- wf %>% add_frosting(f)
p <- forecast(wf1)
p
#> An `epi_df` object, 21 x 4 with metadata:
#> * geo_type = state
#> * time_type = day
#> * as_of = 2023-03-10
#>
#> # A tibble: 21 × 4
#> geo_value time_value ahead distn
#> <chr> <date> <int> <qtls(7)>
#> 1 ak 2021-12-31 1 [0.0399]
#> 2 ak 2021-12-31 2 [0.0571]
#> 3 ak 2021-12-31 3 [0.0656]
#> 4 ak 2021-12-31 4 [0.0821]
#> 5 ak 2021-12-31 5 [0.0881]
#> 6 ak 2021-12-31 6 [0.0865]
#> 7 ak 2021-12-31 7 [0.0549]
#> 8 ca 2021-12-31 1 [0.14]
#> 9 ca 2021-12-31 2 [0.148]
#> 10 ca 2021-12-31 3 [0.155]
#> # ℹ 11 more rows