layer_residual_quantiles.Rd

% Generated by roxygen2: do not edit by hand
% Please edit documentation in R/layer_residual_quantiles.R
\name{layer_residual_quantiles}
\alias{layer_residual_quantiles}
\title{Creates predictions based on residual quantiles}
\usage{
layer_residual_quantiles(
  frosting,
  ...,
  quantile_levels = c(0.05, 0.1, 0.25, 0.5, 0.75, 0.9, 0.95),
  symmetrize = TRUE,
  by_key = character(0L),
  name = ".pred_distn",
  id = rand_id("residual_quantiles")
)
}
\arguments{
\item{frosting}{a \code{frosting} postprocessor}

\item{...}{Unused, include for consistency with other layers.}

\item{quantile_levels}{numeric vector of probabilities with values in (0,1)
referring to the desired quantile. Note that 0.5 will always be included
even if left out by the user.}

\item{symmetrize}{logical. If \code{TRUE} then interval will be symmetric.}

\item{by_key}{A character vector of keys to group the residuals by before
calculating quantiles. The default, \code{c()} performs no grouping.}

\item{name}{character. The name for the output column.}

\item{id}{a random id string}
}
\value{
an updated \code{frosting} postprocessor with additional columns of the
residual quantiles added to the prediction
}
\description{
Creates predictions based on residual quantiles
}
\examples{
library(dplyr)
jhu <- covid_case_death_rates \%>\%
  filter(time_value > "2021-11-01", geo_value \%in\% c("ak", "ca", "ny"))

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_residual_quantiles(
    quantile_levels = c(0.025, 0.975),
    symmetrize = FALSE
  ) \%>\%
  layer_naomit(.pred)
wf1 <- wf \%>\% add_frosting(f)

p <- forecast(wf1)

f2 <- frosting() \%>\%
  layer_predict() \%>\%
  layer_residual_quantiles(
    quantile_levels = c(0.3, 0.7),
    by_key = "geo_value"
  ) \%>\%
  layer_naomit(.pred)
wf2 <- wf \%>\% add_frosting(f2)

p2 <- forecast(wf2)
}