You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We currently only support adding supplemental attributes one at a time. This is likely to be slow if there are thousands of attributes because each addition causes one row to be added to an SQLite table. We support bulk addition of time series for the same reason and could do the same for supplemental attributes.
The text was updated successfully, but these errors were encountered:
I made a prototype. Performance is ~3x faster with bulk additions. My benchmark adds 500k supplemental attributes to components.
add 1 at a time: 24.378 seconds
add 1000 at a time: 7.7 seconds
The code is straightforward and I'm not worried about that. There is a cost in user complexity for the bulk additions. The overall time involved here is not huge. Is it sufficient to justify the complexity (if users notice the docstrings)?
We currently only support adding supplemental attributes one at a time. This is likely to be slow if there are thousands of attributes because each addition causes one row to be added to an SQLite table. We support bulk addition of time series for the same reason and could do the same for supplemental attributes.
The text was updated successfully, but these errors were encountered: