Documentation for MySejahtera datasets

Note: As per the MySejahtera privacy policy, individual-level check-in data is purged after 90 days. These summary statistics are stored only as aggregated totals; MySejahtera does not store the underlying data. Consequently, data revisions are not possible for dates more than 90 days ago, even if an inconsistency is spotted.

File naming convention

1. checkin_malaysia.csv: Static name; file is updated by 1500hrs daily
2. checkin_malaysia_time.csv: Static name; file is updated by 1500hrs daily
3. trace_malaysia.csv: Static name; file is updated by 1500hrs daily

Variables and Methodology

1. date: yyyy-mm-dd format; data correct as of 2359hrs on that date
2. checkins: number of checkins at all locations registered on MySejahtera
3. unique_ind: number of unique accounts which checked in
4. unique_loc: number of unique premises checked into
5. i: in the time density file, checkins are aggregated by half-hour buckets, giving 48 in total; bucket i corresponds to the ith half-hour slot of the day. for instance, i = 0 corresponds to 0000 - 0029; i = 31 corresponds to 1500 - 1529.
6. casual_contacts: number of casual contacts identified and notified by CPRC's automated contact tracing system
7. hide_large: number of large hotspots identified by CPRC's hotspot identification system
8. hide_small: number of small hotspots identified by CPRC's hotspot identification system