Hello there

The primary reason why Terra daily, body, nutrition and menstruation payloads contain all readings from the beginning of the day is to account for updates to readings that a provider may perform after the reading is taken, for example: Garmin may update scores after a period of activity. Additionally, this reflects how most data appears users in the app or web-portal provided by the device manufacturer (split by day).

In order to deduplicate Terra payloads:

• Identify the start_time found in the metadata object in each data item and overrite the file or record in your database or bucket, with the new payload (or override if you wish to keep older payloads). If you make a REST request that spans several days, data will be chunked by date (one data object per day)
• For activity and sleep payloads, deduplication and chunking is different as a sleep/workout session can span across days. Each data item for a sleep/activity payload records one event i.e a nap, sleep, workout, run, exercise, match etc. To deduplicate these, use the summary_id property available in the metadata object to consolodate the same events.