Skip to contents

Modify Reference Date

Source: R/modify_reference_date.R
modify_reference_date.Rd

modify_reference_date applies an arbitrary transformation to the reference_date of a hubverse table and recomputes the horizon, based on the new reference_date and existing target_end_date. A new unit for the horizon can also be specified.

Usage

modify_reference_date(
  original_hub_tbl,
  horizon_timescale = NULL,
  reference_date_transform = identity,
  reference_date_col = "reference_date",
  target_end_date_col = "target_end_date",
  horizon_col = "horizon",
  horizon_timescale_col = "horizon_timescale"
)

Arguments

original_hub_tbl

A data frame containing the hub table.

horizon_timescale

The timescale for the horizon. Must be a valid argument to horizons_from_target_end_dates(). If NULL, the horizon from horizon_timescale_col is used.

reference_date_transform

A function to transform the reference date. Default is identity.

reference_date_col

The name of the reference date column. Default is "reference_date".

target_end_date_col

The name of the target end date column. Default is "target_end_date".

horizon_col

The name of the horizon column. Default is "horizon".

horizon_timescale_col

The name of the horizon timescale column. Default is "horizon_timescale".

Value

A modified data frame with the reference date transformed and horizons recomputed

Examples

example_daily_forecast_flu |>
  trajectories_to_quantiles(
    id_cols = "location",
    timepoint_col = "date",
    value_col = "hosp"
  ) |>
  dplyr::rename(target_end_date = date) |>
  get_hubverse_quantile_table(
    reference_date = as.Date("2023-10-20"),
    horizon_timescale = "weeks",
    target_name = "my_target",
    timepoint_col = "target_end_date",
    horizons = 1:4
  ) |>
  modify_reference_date(\(x) x - 1, horizon_timescale = "days")
#> # A tibble: 4,876 × 9
#>    reference_date target    horizon horizon_timescale target_end_date location
#>    <date>         <chr>       <dbl> <chr>             <date>          <chr>   
#>  1 2023-10-19     my_target       8 days              2023-10-27      AK      
#>  2 2023-10-19     my_target       8 days              2023-10-27      AK      
#>  3 2023-10-19     my_target       8 days              2023-10-27      AK      
#>  4 2023-10-19     my_target       8 days              2023-10-27      AK      
#>  5 2023-10-19     my_target       8 days              2023-10-27      AK      
#>  6 2023-10-19     my_target       8 days              2023-10-27      AK      
#>  7 2023-10-19     my_target       8 days              2023-10-27      AK      
#>  8 2023-10-19     my_target       8 days              2023-10-27      AK      
#>  9 2023-10-19     my_target       8 days              2023-10-27      AK      
#> 10 2023-10-19     my_target       8 days              2023-10-27      AK      
#> # ℹ 4,866 more rows
#> # ℹ 3 more variables: output_type <chr>, output_type_id <dbl>, value <dbl>