Filter observations to a maximum delay and then extract the latest observations. This is useful for model evaluation where you want to assess performance against the data as the model would have seen it.
Arguments
- obs
A
data.framecontaining at least the following variables:reference date(index date of interest),report_date(report date for observations), andconfirm(cumulative observations by reference and report date).- max_delay
The maximum delay to model in the delay distribution, specified in units of the timestep (e.g., if
timestep = "week", thenmax_delay = 3means 3 weeks). If not specified the maximum observed delay is assumed to be the true maximum delay in the model. Otherwise, an integer greater than or equal to 1 can be specified. Observations with delays larger than the maximum delay will be dropped. If the specified maximum delay is too short, nowcasts can be biased as important parts of the true delay distribution are cut off. At the same time, computational cost scales non-linearly with this setting, so you want the maximum delay to be as long as necessary, but not much longer.Steps to take to determine the maximum delay:
Consider what is realistic and relevant for your application.
Check the proportion of observations reported (
prop_reported) by delay in thenew_confirmoutput ofenw_preprocess_obs.Use
check_max_delay()to check the coverage of a candidatemax_delay.If in doubt, check if increasing the maximum delay noticeably changes the delay distribution or nowcasts as estimated by
epinowcast. If it does, your maximum delay may still be too short.
Note that delays are zero indexed and so include the reference date and
max_delay - 1other intervals (i.e. amax_delayof 1 corresponds to no delay).- timestep
The timestep to used in the process model (i.e. the reference date model). This can be a string ("day", "week", "month") or a numeric whole number representing the number of days. If your data does not have this timestep then you may wish to make use of
enw_aggregate_cumulative()to aggregate your data to the desired timestep.
Value
A data.table of observations filtered for the
latest available data for each reference date at the
specified maximum delay.
See also
Preprocessing functions
enw_add_delay(),
enw_add_max_reported(),
enw_add_metaobs_features(),
enw_assign_group(),
enw_complete_dates(),
enw_construct_data(),
enw_extend_date(),
enw_filter_delay(),
enw_filter_reference_dates(),
enw_filter_report_dates(),
enw_flag_observed_observations(),
enw_impute_na_observations(),
enw_latest_data(),
enw_metadata(),
enw_metadata_delay(),
enw_missing_reference(),
enw_preprocess_data(),
enw_reporting_triangle(),
enw_reporting_triangle_to_long()
Examples
obs <- enw_example("preprocessed")$obs[[1]]
enw_obs_at_delay(obs, max_delay = 2)
#> reference_date .group report_date max_confirm location age_group confirm
#> <IDat> <num> <IDat> <int> <fctr> <fctr> <int>
#> 1: 2021-07-14 1 2021-07-15 72 DE 00+ 34
#> 2: 2021-07-15 1 2021-07-16 69 DE 00+ 43
#> 3: 2021-07-16 1 2021-07-17 47 DE 00+ 32
#> 4: 2021-07-17 1 2021-07-18 65 DE 00+ 27
#> 5: 2021-07-18 1 2021-07-19 50 DE 00+ 15
#> 6: 2021-07-19 1 2021-07-20 36 DE 00+ 19
#> 7: 2021-07-20 1 2021-07-21 94 DE 00+ 55
#> 8: 2021-07-21 1 2021-07-22 91 DE 00+ 46
#> 9: 2021-07-22 1 2021-07-23 99 DE 00+ 53
#> 10: 2021-07-23 1 2021-07-24 86 DE 00+ 42
#> 11: 2021-07-24 1 2021-07-25 93 DE 00+ 39
#> 12: 2021-07-25 1 2021-07-26 74 DE 00+ 12
#> 13: 2021-07-26 1 2021-07-27 28 DE 00+ 15
#> 14: 2021-07-27 1 2021-07-28 78 DE 00+ 46
#> 15: 2021-07-28 1 2021-07-29 156 DE 00+ 79
#> 16: 2021-07-29 1 2021-07-30 135 DE 00+ 84
#> 17: 2021-07-30 1 2021-07-31 114 DE 00+ 56
#> 18: 2021-07-31 1 2021-08-01 126 DE 00+ 54
#> 19: 2021-08-01 1 2021-08-02 77 DE 00+ 12
#> 20: 2021-08-02 1 2021-08-03 59 DE 00+ 26
#> 21: 2021-08-03 1 2021-08-04 149 DE 00+ 94
#> 22: 2021-08-04 1 2021-08-05 166 DE 00+ 94
#> 23: 2021-08-05 1 2021-08-06 133 DE 00+ 66
#> 24: 2021-08-06 1 2021-08-07 137 DE 00+ 78
#> 25: 2021-08-07 1 2021-08-08 139 DE 00+ 55
#> 26: 2021-08-08 1 2021-08-09 97 DE 00+ 27
#> 27: 2021-08-09 1 2021-08-10 58 DE 00+ 37
#> 28: 2021-08-10 1 2021-08-11 175 DE 00+ 121
#> 29: 2021-08-11 1 2021-08-12 233 DE 00+ 133
#> 30: 2021-08-12 1 2021-08-13 237 DE 00+ 137
#> 31: 2021-08-13 1 2021-08-14 204 DE 00+ 130
#> 32: 2021-08-14 1 2021-08-15 189 DE 00+ 115
#> 33: 2021-08-15 1 2021-08-16 125 DE 00+ 46
#> 34: 2021-08-16 1 2021-08-17 98 DE 00+ 55
#> 35: 2021-08-17 1 2021-08-18 242 DE 00+ 181
#> 36: 2021-08-18 1 2021-08-19 223 DE 00+ 178
#> 37: 2021-08-19 1 2021-08-20 202 DE 00+ 171
#> 38: 2021-08-20 1 2021-08-21 171 DE 00+ 159
#> 39: 2021-08-21 1 2021-08-22 112 DE 00+ 112
#> 40: 2021-08-22 1 2021-08-22 45 DE 00+ 45
#> reference_date .group report_date max_confirm location age_group confirm
#> <IDat> <num> <IDat> <int> <fctr> <fctr> <int>
#> cum_prop_reported delay
#> <num> <num>
#> 1: 0.4722222 1
#> 2: 0.6231884 1
#> 3: 0.6808511 1
#> 4: 0.4153846 1
#> 5: 0.3000000 1
#> 6: 0.5277778 1
#> 7: 0.5851064 1
#> 8: 0.5054945 1
#> 9: 0.5353535 1
#> 10: 0.4883721 1
#> 11: 0.4193548 1
#> 12: 0.1621622 1
#> 13: 0.5357143 1
#> 14: 0.5897436 1
#> 15: 0.5064103 1
#> 16: 0.6222222 1
#> 17: 0.4912281 1
#> 18: 0.4285714 1
#> 19: 0.1558442 1
#> 20: 0.4406780 1
#> 21: 0.6308725 1
#> 22: 0.5662651 1
#> 23: 0.4962406 1
#> 24: 0.5693431 1
#> 25: 0.3956835 1
#> 26: 0.2783505 1
#> 27: 0.6379310 1
#> 28: 0.6914286 1
#> 29: 0.5708155 1
#> 30: 0.5780591 1
#> 31: 0.6372549 1
#> 32: 0.6084656 1
#> 33: 0.3680000 1
#> 34: 0.5612245 1
#> 35: 0.7479339 1
#> 36: 0.7982063 1
#> 37: 0.8465347 1
#> 38: 0.9298246 1
#> 39: 1.0000000 1
#> 40: 1.0000000 0
#> cum_prop_reported delay
#> <num> <num>