cdc_baseline_args_list: CDC baseline forecaster argument constructor

In cmu-delphi/epipredict: Basic epidemiology forecasting methods

CDC baseline forecaster argument constructor

Description

Constructs a list of arguments for cdc_baseline_forecaster().

Usage

cdc_baseline_args_list(
  data_frequency = "1 week",
  aheads = 1:5,
  n_training = Inf,
  forecast_date = NULL,
  quantile_levels = c(0.01, 0.025, 1:19/20, 0.975, 0.99),
  nsims = 100000L,
  symmetrize = TRUE,
  nonneg = TRUE,
  quantile_by_key = "geo_value",
  ...
)

Arguments

data_frequency	Integer or string. This describes the frequency of the input epi_df. For typical FluSight forecasts, this would be "1 week". Allowable arguments are integers (taken to mean numbers of days) or a string like "7 days" or "2 weeks". Currently, all other periods (other than days or weeks) result in an error.
aheads	Integer vector. Unlike arx_forecaster(), this doesn't have any effect on the predicted values. Predictions are always the most recent observation. This determines the set of prediction horizons for layer_cdc_flatline_quantiles()⁠. It interacts with the ⁠data_frequency⁠argument. So, for example, if the data is daily and you want forecasts for 1:4 days ahead, then you would use⁠1:4⁠. However, if you want one-week predictions, you would set this as ⁠c(7, 14, 21, 28)⁠. But if ⁠data_frequencyis"1 week"⁠, then you would set it as ⁠1:4'.
n_training	Integer. An upper limit for the number of rows per key that are used for training (in the time unit of the epi_df).
forecast_date	Date. The date from which the forecast is occurring. The default NULL will determine this automatically from either the maximum time value for which there's data if there is no latency adjustment (the default case), or the as_of date of epi_data if adjust_latency is non-NULL.
quantile_levels	Vector or NULL. A vector of probabilities to produce prediction intervals. These are created by computing the quantiles of training residuals. A NULL value will result in point forecasts only.
nsims	Positive integer. The number of draws from the empirical CDF. These samples are spaced evenly on the (0, 1) scale, F_X(x) resulting in linear interpolation on the X scale. This is achieved with stats::quantile() Type 7 (the default for that function).
symmetrize	Logical. The default TRUE calculates symmetric prediction intervals. This argument only applies when residual quantiles are used. It is not applicable with trainer = quantile_reg(), for example.
nonneg	Logical. Force all predictive intervals be non-negative. Because non-negativity is forced before propagating forward, this has slightly different behaviour than would occur if using layer_threshold().
quantile_by_key	Character vector. Groups residuals by listed keys before calculating residual quantiles. See the by_key argument to layer_residual_quantiles() for more information. The default, character(0) performs no grouping. This argument only applies when residual quantiles are used. It is not applicable with trainer = quantile_reg(), for example.
...	Space to handle future expansions (unused).

Value

A list containing updated parameter choices with class cdc_flat_fcast.

Examples

cdc_baseline_args_list()
cdc_baseline_args_list(symmetrize = FALSE)
cdc_baseline_args_list(quantile_levels = c(.1, .3, .7, .9), n_training = 120)

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