Open sandorkertesz opened 5 months ago
All modified and coverable lines are covered by tests :white_check_mark:
Project coverage is 94.57%. Comparing base (
3aa638b
) to head (f51d06d
).
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
This PR adds an in-memory LRU cache for the interpolation matrices. The cache has a memory limit which is set to 300 MB by default. The memory limit can also be set to 0 (no caching) or None (no upper memory limit).
The memory limit configuration is not yet exposed to the users (hardcoded).
With this cache repeated interpolations between the same input and output grids can be run much faster. The speed-up is around x10 on a MacBook using the "largest" interpolation available at the moment: