release 0.6.0
Since 0.5.0 we fixed the following issues:
- #66 (closed): (refactor) learner.tell(x, None) might be renamed to learner.tell_pending(x)
- #92 (closed): DeprecationWarning: sorted_dict.iloc is deprecated. Use SortedDict.keys() instead.
- #94 (closed): Learner1D breaks if right bound is added before the left bound
- #95 (closed): Learner1D's bound check algo in self.ask doesn't take self.data or self.pending_points
- #96 (closed): Learner1D fails when function returns a list instead of a numpy.array
- #97 (closed): Learner1D fails when a point (x, None) is added when x already exists
- #98 (closed): Learner1D.ask breaks when adding points in some order
- #99 (closed): Learner1D doesn't correctly set the interpolated loss when a point is added
- #101 (closed): How should learners handle data that is outside of the domain
- #102 (closed): No tests for the 'BalancingLearner'
- #105 (closed): LearnerND fails for BalancingLearner test
- #108 (closed): (BalancingLearner) loss is cached incorrectly
- #109 (closed): Learner2D suggests same point twice
and merged the following Merge Requests:
- !93 (merged): add a release guide
- !94 (merged): add runner.max_retries
- !95 (merged): 1D: fix the rare case where the right boundary point exists before the left bound
- !96 (merged): More efficient 'tell_many'
- !97 (merged): Fix #97 (closed) and #98 (closed)
- !98 (merged): Resolve "DeprecationWarning: sorted_dict.iloc is deprecated. Use SortedDict.keys() instead."
- !99 (merged): Resolve "Learner1D's bound check algo in self.ask doesn't take self.data or self.pending_points"
- !100 (merged): Resolve "Learner1D doesn't correctly set the interpolated loss when a point is added"
- !101 (merged): Resolve "Learner1D fails when function returns a list instead of a numpy.array"
- !102 (merged): introduce 'runner.retries' and 'runner.raise_if_retries_exceeded'
- !103 (merged): 2D: rename 'learner._interp' to 'learner.pending_points' as in other learners
- !104 (merged): Make the AverageLearner only return new points ...
- !105 (merged): move specific tests for a particular learner to separate files
- !107 (merged): Introduce 'tell_pending' which replaces 'tell(x, None)'
- !112 (merged): Resolve "LearnerND fails for BalancingLearner test"
- !113 (merged): Resolve "(BalancingLearner) loss is cached incorrectly"
- !114 (merged): update release guide to add a 'dev' tag on top of regular tags
- !115 (merged): Resolve "How should learners handle data that is outside of the domain"
- !116 (merged): 2D: fix #109 (closed)
New features
- add
learner.tell_pendingwhich replaceslearner.tell(x, None) - add error-handling with
runner.retriesandrunner.raise_if_retries_exceeded - make
learner.pending_pointsandrunner.pending_pointspublic API - rename
learner.ask(n, add_data)->learner.ask(n, tell_pending) - added the
overheadmethod to theBlockingRunner
I think we should release 0.6.0 since we introduce some new features. @jbweston and @anton-akhmerov do you have a different opinion?
Activity
assigned to @anton-akhmerov
assigned to @jbweston and unassigned @anton-akhmerov
unassigned @jbweston
I would like the following to be merged as well: !112 (merged) #105 (closed)
Since this issue is actually much broader than just the balancingLearner + LearnerND, the problem is actually the combination of Async + LearnerND, which I suppose would occur often.
Then maybe we can also add !113 (merged) (#108 (closed))
@Jorn tag any issues and MRs with the v0.6 milestone (in the right-hand menu in Gitlab), i.e. please triage the issues and MRs (I said I would do this above, but you're probably better placed to do it).
Edited by Joseph WestonAll issues and merge requests for 0.6 have been closed and merged. Shall we do it now? @anton-akhmerov, @Jorn, and @jbweston?
