Changelog
---
* `dimod` 0.9.4 -> 0.9.5
- `BQM.normalize` now returns the value the BQM was scaled by
- `SampleSet.relabel_variables` no longer blocks for unresolved sample sets
- `FileView` has a new parameter, `ignore_variables` that treats the BQM as integer-labelled
- `ScaleComposite` no longer blocks
- `FileView.seek` now works correctly with `io.SEEK_END`
- Change: `BinaryPolynomial` support in `ScaleComposite` has been removed
* `dwave-cloud-client` 0.7.4 -> 0.7.5
- User-Agent extended with platform-specific tags
- CLI: `dwave --platform`
- CLI: additive verbosity for install/setup commands
- Fixed: `order_by` argument to `Client.get_solver` affected effective solver
filters if used in isolation (401)
- Fixed: during problem multipart upload, individual part upload could fail on
retry attempt if part data stream was exhausted in previous upload attempt
* `dwave-hybrid` 0.5.0 -> 0.5.1
- All QPU samplers now use client-based QPU selection instead of solver feature-based,
therefore allowing users to define the solver via Ocean configuration
* `dwave-system` 0.9.6 -> 0.9.9
- Added `DWaveCliqueSampler`
- Added `order_by` kwarg in `DWaveSampler` constructor
- `LeapHybridSampler` uses BQM serialization version 2.0
- Embedding composites no longer block
- `MockDWaveSampler` now has the correct topology
- Fix rounding in anneal schedule validation
* `minorminer` 0.1.9 -> 0.2.0
- Added polytime clique embedder for chimera and pegasus graphs under `minorminer.busclique`
(an implementation to compute biclique embeddings is also available but needs improvement)
- Added layout and placement strategies under `minorminer.layout`
- Improved restart behavior (controlled through `tries` parameter)
Package versions
---
dimod==0.9.5
dwave-cloud-client==0.7.5
dwave-hybrid==0.5.1
dwave-inspector==0.2.2
dwave-neal==0.5.4
dwave-networkx==0.8.6
dwave-qbsolv==0.3.0
dwave-system==0.9.9
dwave-tabu==0.3.0
dwavebinarycsp==0.1.2
minorminer==0.2.0
penaltymodel==0.16.3
penaltymodel-cache==0.4.1
penaltymodel-lp==0.1.3
pyqubo>=0.4.0
Extras, `all`
---
penaltymodel-mip==0.2.3
penaltymodel-maxgap==0.5.3