Construct a csfmt_ensemble_v3
Usage
csfmt_ensemble_v3(data, id_cols, time_col = "isoyearweek", draws = list())Arguments
- data
data.table with the identity columns and `time_col`.
- id_cols
Character vector of identity columns defining a series.
- time_col
Time-ordering column (default "isoyearweek").
- draws
Optional named list of `[nrow(data) x n_draws]` matrices, given in `data`'s input row order (they are reordered to match the canonical sort).
See also
vignette("pipeline", package = "csalert") is built on this
format: its nowcast engine produces one and ens_collapse
reduces it. The vignette never calls this constructor directly, because the
engines build the ensemble for you. Call it yourself only when you already
hold draws from somewhere else.
Other ensemble format functions:
print.csfmt_ensemble_v3(),
set_time_series_id(),
validate_ensemble()
Examples
d <- data.table::data.table(
location_code = "nation",
age = "total",
isoyearweek = c("2023-01", "2023-02", "2023-03")
)
set.seed(1)
ens <- csfmt_ensemble_v3(
d,
id_cols = c("location_code", "age"),
draws = list(numerator_nowcasted = matrix(rpois(3 * 100, 20), nrow = 3))
)
ens
#> <csfmt_ensemble_v3> 3 rows | 1 series | draws: numerator_nowcasted
# $data carries the identity + the canonical sort keys. The trailing []
# forces the print: data.table suppresses the first auto-print of a table
# that was last modified by reference, which the constructor does.
ens$data[]
#> Key: <time_series_id, time_series_internal_id>
#> location_code age isoyearweek time_series_id time_series_label
#> <char> <char> <char> <char> <char>
#> 1: nation total 2023-01 0ec581372be7c933 nation\037total
#> 2: nation total 2023-02 0ec581372be7c933 nation\037total
#> 3: nation total 2023-03 0ec581372be7c933 nation\037total
#> time_series_internal_id
#> <int>
#> 1: 1
#> 2: 2
#> 3: 3
# $draws holds one [weeks x draws] matrix per measure
dim(ens$draws$numerator_nowcasted)
#> [1] 3 100