Aggregates mortality data to a weekly basis. Where the percentiles and number of mortalities obtained after every week up to n_week is also given. For more details see the help vignette:
Arguments
- data
Dataset containing doe (Date of event), dor (Date of registation) and location_code. The columns must have these exact names.
- aggregation_date
Date of aggregation
- n_week
Number of weeks to calculate the percentage of the total registraations. Must be larger og equal to 2 amd smaller than the total number of weeks in the dataset.
- pop_data
Population data, must contain a column called pop with the population data and a column with year and possibly week.
Examples
data <- attrib::data_fake_nowcasting_nation_raw
data[doe < as.Date("2019-01-01")]
#> doe dor location_code
#> 1: 2018-01-01 2018-01-18 nation_nor
#> 2: 2018-01-01 2018-01-27 nation_nor
#> 3: 2018-01-01 2018-01-22 nation_nor
#> 4: 2018-01-01 2018-01-16 nation_nor
#> 5: 2018-01-01 2018-01-29 nation_nor
#> ---
#> 32851: 2018-12-31 2019-01-19 nation_nor
#> 32852: 2018-12-31 2019-01-14 nation_nor
#> 32853: 2018-12-31 2019-01-30 nation_nor
#> 32854: 2018-12-31 2019-01-28 nation_nor
#> 32855: 2018-12-31 2019-01-21 nation_nor
aggregation_date <- as.Date("2020-01-01")
n_week <- 3
data_aggregated <- nowcast_aggregate(data, aggregation_date, n_week)