Closed wilsonbb closed 4 months ago
Check out this pull request on
See visual diffs & provide feedback on Jupyter Notebooks.
Powered by ReviewNB
Before [89aa1169] | After [49c8c480] | Ratio | Benchmark (Parameter) |
---|---|---|---|
33.7±0.2ms | 34.2±0.3ms | 1.02 | benchmarks.time_prune_sync_workflow |
33.6±0.3ms | 33.2±0.4ms | 0.99 | benchmarks.time_batch |
Click here to view all benchmarks.
Attention: Patch coverage is 77.77778%
with 2 lines
in your changes are missing coverage. Please review.
Project coverage is 95.82%. Comparing base (
89aa116
) to head (95fb692
).
Files | Patch % | Lines |
---|---|---|
src/tape/ensemble.py | 77.77% | 2 Missing :warning: |
:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.
Thank you for looking into this. Did you try playing with the dask version itself at all? Their test suite is failing atm, and I wonder if pinned dask itself to something older would stabilize things
I've messed around a bit, but among the compatible dask versions, I didn't see the error disappear. But probably worth checking again
Reverts lincc-frameworks/tape#420 and pins dask-expr <1.0.11 to address flakiness reported in https://github.com/lincc-frameworks/tape/issues/434. The plan is to do another PR for the batch single_lc changes in #420 after all of the batch-related flakiness has been addressed.
Note that even with these changes we're still seeing the test_batch_by_band errors referenced in lincc-frameworks/tape#434 consistently on the python 3.11.9 build, but this is still an improvement over the current test failures we've been seeing in https://github.com/lincc-frameworks/tape/issues/427