Open 7yl4r opened 1 year ago
Based on the size of the data I would image a disk cache, right?
Yes, I was imagining:
To save:
{param_hash}_{record_count}.p
) tempfile packageto fetch before performing query:
I'm not certain if record_count is the only parameter to look at, however. Records could be updated without changing the count of records.
I usually go to https://joblib.readthedocs.io/en/latest/auto_examples/memory_basic_usage.html for that kind of operation. If you choose to not save on disk we can use functools.lru_cache
from the standard library.
a nice feature to have would be to allow retention of cached data so that data does not need to be redownloaded.