Closed smnorris closed 2 years ago
Even if one considers all provided (and non-internally generated) columns in the source table, the data can have duplicates. This script should probably sift through the data to generate unique records and not bother storing the observation_id value in the output tables.
Example duplicate:
I think noting the id instability in the README is fine, no need to strip out the id values from the output products. The instability makes it harder to refer back to the source data but is not particularly important for other uses.
It seems the IDs can change over time (as with FISS Obstacles).
There is not much we can do about this but it is worth noting in the README.