Tidy a nowcast into the cross-package nowcast table

Returns the posterior nowcast as one row per event date per stratum, in the column layout shared by every engine tbl.now normalises. Called on a fitted nowcast() it first draws the posterior predictive via predict(); called on the result of predict() it summarises the draws it already holds.

Arguments

x

A nowcast (from nowcast()) or a nowcast_prediction (from predict() on one).

probs

Optional numeric probabilities in [0, 1]. Each adds one column named q<probs * 100> (so 0.05 gives q5, 0.025 gives q2.5). Because the object keeps the draws, these quantiles are exact.

conf.level

Width of the credible interval (default 0.95).

...

Passed to predict() when x is a fitted nowcast (e.g. n_draws, seed); unused otherwise.

Value

A tibble::tibble() sorted by stratum then event_date, with columns event_date (Date), stratum (character, "all" when the fit is unstratified), estimate (posterior median), conf.low, conf.high, level (the width the interval actually has) and engine, plus one column per entry of probs.

See also

model_parameters() for the per-parameter table that tidy() used to return.

Examples

if (requireNamespace("tbl.now", quietly = TRUE)) {
  # nc <- nowcast(data, model(nb_likelihood(), hsgp_epidemic(), lognormal_delay()))
  # tidy(nc)                                    # one row per event date x stratum
  # tidy(predict(nc), probs = c(0.05, 0.95))    # plus exact q5 / q95 columns
  # model_parameters(nc)                        # one row per estimated parameter
}
#> NULL