Tries to impute missing values by performing smart assignment on all columns that are missing data.
E.g. if location_code='norge'
then we know that granularity_geo='nation'
.
Arguments
- x
An object of type nowcast_aggregate_data_v1 created by
nowcast_aggregate
- ...
Not used.
nowcast_aggregate_data_v1
The **columns in bold** will be used to impute the listed columns.
**location_code**: - granularity_geo - country_iso3
**isoyear** (when `granularity_time=="isoyear"`): - isoweek - isoyearweek - season - seasonweek - calyear - calmonth - calyearmonth - date
**isoyearweek** (when `granularity_time=="isoweek"`): - isoyear - isoweek - season - seasonweek - calyear - calmonth - calyearmonth - date
**date** (when `granularity_time=="day"`): - isoyear - isoweek - isoyearweek - season - seasonweek - calyear - calmonth - calyearmonth
With regards to the time columns, `granularity_time` takes precedence over everything. If `granularity_time` is missing, then we try to impute `granularity_time` by seeing if there is only one time column with non-missing data. Due to the multitude of time columns, `granularity_time` is an extremely important column and should always be kept with valid values.