layer_threshold: Lower and upper thresholds for predicted values

In cmu-delphi/epipredict: Basic epidemiology forecasting methods

Lower and upper thresholds for predicted values

Description

This postprocessing step is used to set prediction values that are smaller than the lower threshold or higher than the upper threshold equal to the threshold values.

Usage

layer_threshold(
  frosting,
  ...,
  lower = 0,
  upper = Inf,
  id = rand_id("threshold")
)

Arguments

frosting	a frosting postprocessor
...	<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 like x:y can be used to select a range of variables. Typical usage is .pred to threshold predictions to a range (say, nonnegative).
lower	Lower threshold for the prediction values. That is, any predictions that are less than this lower bound are set to it. Default value is 0.
upper	Upper threshold for the prediction values. That is, any predictions that are greater than this upper bound are set to it. Default value is Inf.
id	a random id string

Value

an updated frosting postprocessor

Examples

jhu <- covid_case_death_rates %>%
  filter(time_value < "2021-03-08", geo_value %in% c("ak", "ca", "ar"))

r <- epi_recipe(jhu) %>%
  step_epi_lag(death_rate, lag = c(0, 7, 14)) %>%
  step_epi_ahead(death_rate, ahead = 7) %>%
  step_epi_naomit()
wf <- epi_workflow(r, linear_reg()) %>% fit(jhu)

f <- frosting() %>%
  layer_predict() %>%
  layer_threshold(.pred, lower = 0.180, upper = 0.310)
wf <- wf %>% add_frosting(f)
p <- forecast(wf)
p

cmu-delphi/epipredict documentation built on March 5, 2025, 12:17 p.m.