FIX reproducibility and parallelization of InstanceHardnessThreshold #598
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This PR aims to solve couple problems with existing
InstanceHardnessThreshold
sampler.estimator
is notNone
, result won't not be reproducible ifestimator
doesn't haverandom_state
.estimator
is notNone
, it may have differentn_jobs
value than one given toInstanceHardnessThreshold
constructor. So when givenestimator
'sn_jobs
equals 1, settingn_jobs>1
inInstanceHardnessThreshold
won't affect anything, andfit_resample
will run in single thread.n_jobs
in both cases match, by moving parallelism away fromestimator
tocross_val_predict
, this enables coarse-grained parallelism, possibly speeding up computation.