Open anugrah-saxena opened 3 years ago
The default should be sample date when available, but received date if the sample data is unavailable. This field should be retroactively filled.
This is risky @mazeller Sample date is the date that the sample was collected. The Received date is not linked to the sampling event and could be off enough to impact inference (though, generally, only a few days based on what I see in the USDA data). Received date should only be used if absolutely necessary, and shouldn't be used to the day for Bayesian analyses (in these case, the YYYY-MM would have confidence [unless the received date is at beginning/end]). So - internally - yes, but for these data it should not default to "received date" for people who are doing analysis. Does that make sense? OK for plotting on FLUture, but definitely not for phylodynamic analyses.
Received date should be updated and equal to either the sample date or the received date for all data
@mazeller can you check the requirements again and update here?