Closed hamishgibbs closed 3 years ago
This should be a relatively simple set difference. We should make sure to retain the original who_ids
and who_code
for any problem IDs.
@todowede - I will get started on this one
From CG: There should be no values in the previous_measure_number and following_measure_number for who_codes 10, 11, 12, 13 if there are they can simply be removed.
This is added in #141.
We need to produce summary tables of IDs that are recorded in ID columns (
previous_measure_number
,following_measure_number
) but are not found in the who_id field.Most of these will be from the very early days of the dataset.
For records that are not included in the release data, i.e. records with
who_code
in["10", "11", "12", "13"]
- id problems are less of an issue.It is particularly important that we identify any ID problems in records that are being included in release data.