Dwave-ocean-sdk

Changelog
---

* `dwave-cloud-client` 0.8.2 -> 0.8.4

- All SAPI requests retried whenever possible, controlled with new `http_retry_*` Client options (414)
- Introduced `SAPIError` base exception with SAPI error message and code, other exceptions reorganized in a backwards-compatible way
- Support for problem labels: `sample_*` methods accept optional `label` kwarg, and `Future.sampleset.info` contains `problem_label` (444)
- Added general problem info (id/label) to unstructured solvers' response (449)
- Switched to CircleCI (119)
- Handle remote disconnects in an edge case (229)

* `dwave-inspector` 0.2.4 -> 0.2.5

- Problem label field under problem details dropdown
- Support for Python 3.9

* `dwave-system` 1.3.0 -> 1.4.0

- Support for problem labels (all samplers now accept an optional `label` kwarg)
- Label on `VirtualGraphComposite` flux bias measurements
- Hybrid samplers now properly select the latest version of an appropriate hybrid solver
- Python 2 cruft cleanup

* `pyqubo` 1.0.7 -> 1.0.10

- Support for Python 3.9


Package versions
---

dimod==0.9.13
dwave-cloud-client==0.8.4
dwave-greedy==0.1.2
dwave-hybrid==0.6.1
dwave-inspector==0.2.5
dwave-neal==0.5.7
dwave-networkx==0.8.8
dwave-qbsolv==0.3.2
dwave-system==1.4.0
dwave-tabu==0.3.1
dwavebinarycsp==0.1.2
minorminer==0.2.5
penaltymodel-cache==0.4.3
penaltymodel-lp==0.1.4
penaltymodel==0.16.4
pyqubo==1.0.10

Extras, `all`
---

penaltymodel-maxgap==0.5.4
penaltymodel-mip==0.2.4

Changelog
---

* `dimod` 0.9.10 -> 0.9.13

- `append_variables` function, replacing the funcionality of `SampleSet.append_variables`
- `append_vectors` function, to add more data vactors to a sample set
- Documentation improvements
- Support for python 3.9
- Improved error messages and variable names in `DQM.from_numpy_vectors`
- `.clang-format` file for C++ formatting
- C++ `AdjVectorBQM` and `AdjMapBQM` can now be constructed from COO-formatted arrays
- Python `AdjVectorBQM.from_numpy_vectors` and `AdjMapBQM.from_numpy_vectors` is now much more performant
- Cython `DiscreteQuadrataticModel` object now uses C++ COO constructors rather than an ad hoc implementation
- `gnp_random_bqm` generator for randomly connected BQMs

- Fix c++ target for windows build
- Fix `DQM.from_numpy_vectors` to no longer segfault when given inconsistent inputs
- `DQM.from_numpy_vectors`, `.set_quadratic` and `.set_quadratic_case` now correctly raise a `ValueError` when the user attempts to create an interaction between cases in a variable
- `BQM.from_numpy_vectors` now maintains the correct variable order
- `AdjDictBQM` biases are no longer 0 dimensional arrays when constructed from a numpy matrix
- Fixed column alignment issue when printing sample sets

- Deprecated `SampleSet.append_variables`

* `dwave-cloud-client` 0.8.1 -> 0.8.2

- Improve error messages for multipart upload failures (433)
- Accept looser file-like for upload (435)
- Support for python 3.9 (442)

* `dwave-greedy` 0.1.1 -> 0.1.2

- Support for python 3.9

* `dwave-hybrid` 0.6.0 -> 0.6.1

- Added `beta_range` option to `(Hybridized)PopulationAnnealing` and `CalculateAnnealingBetaSchedule`
- `ProgressBetaAlongSchedule` now outputs `delta_beta` in addition to `beta`
- Made `ProcessPoolExecutor` availability optional to enable use on systems with process limitations
- Expanded range of supported energies in `EnergyWeightedResampler`
- Linear beta schedule and better defaults in `PopulationAnnealing`
- Tests not included in the package anymore
- Deprecated use of `beta` in `EnergyWeightedResampler`. Use `delta_beta` instead.

* `dwave-neal` 0.5.6 -> 0.5.7

- Support for python 3.9

* `dwave-qbsolv` 0.3.1 -> 0.3.2

- Support for python 3.9

* `dwave-system` 1.2.1 -> 1.3.0

- `ReverseBatchStatesComposite` now inherits from `dimod.Initialized` and by default reverse anneals
- `ReverseAdvanceComposite` now reverse anneals by default with schedule `[[0, 1], [1, 0.35], [9, 0.35], [10, 1]]`
- `EmbeddedStructure` is now deep copy-able
- Embedding composites now raise correct warnings even when chain strengths are provided as a dict
- Many `dwave.embedding` utility functions have been migrated to `minorminer`, though they are still imported into their original namespaces

* `dwave-tabu` 0.3.0 -> 0.3.1

- Support for python 3.9

* `minorminer` 0.2.3 -> 0.2.5

- Incorporate functionality from `dwave.embedding` namespace (170)
- Updated python install instructions
- Support for python 3.9

* `penaltymodel` 0.16.3 -> 0.16.4

- Support for python 3.9

* `penaltymodel-cache` 0.4.1 -> 0.4.3

- Added decorator for isolating penaltymodel cache tests, `isolated_cache`
- Fixed caching of problems with no quadratic/linear biases set
- Support for python 3.9

* `penaltymodel-lp` 0.1.3 -> 0.1.4

- Support for python 3.9

* `penaltymodel-maxgap` 0.5.3 -> 0.5.4

- Support for python 3.9

* `penaltymodel-mip` 0.2.3 -> 0.2.4

- Support for python 3.9


Package versions
---

dimod==0.9.13
dwave-cloud-client==0.8.2
dwave-greedy==0.1.2
dwave-hybrid==0.6.1
dwave-inspector==0.2.4
dwave-neal==0.5.7
dwave-networkx==0.8.8
dwave-qbsolv==0.3.2
dwave-system==1.3.0
dwave-tabu==0.3.1
dwavebinarycsp==0.1.2
minorminer==0.2.5
penaltymodel-cache==0.4.3
penaltymodel-lp==0.1.4
penaltymodel==0.16.4
pyqubo==1.0.7 only available on python < 3.9

Extras, `all`
---

penaltymodel-maxgap==0.5.4
penaltymodel-mip==0.2.4

Changelog
---

* `dimod` 0.9.9 -> 0.9.10
- Fix the behavior of the file-like returned by `DQM.to_file`

Changelog
---

* `dimod` 0.9.7 -> 0.9.9
- Support printing sample sets with discrete variables
- New parallel constructor for c++ AdjMapBQM
- `DQM.to_file` now returns a seekable file-like
- `DQM.from_numpy_vectors` now reserves the quadratic vectors which should reduce memory and speed up construction
- `DQM.from_file` now correctly accepts bytes-like objects
- `DQM.to_file`'s `compressed` kwarg has been deprecated in favor of `compress`

* `dwave-system` 1.1.1 -> 1.2.1
- Added `LeapHybridSampler.min_time_limit` method
- Improved documentation of `LeapHybridDQMSampler`
- More sanity checks on `time_limit` in `LeapHybridDQMSampler`
- `LeapHybridDQMSampler.sample_dqm`'s `compressed` keyword argument has been deprecated in favor of `compress`
- Fixed doctest

Package versions
---

dimod==0.9.9
dwave-cloud-client==0.8.1
dwave-greedy==0.1.1
dwave-hybrid==0.6.0
dwave-inspector==0.2.4
dwave-neal==0.5.6
dwave-networkx==0.8.8
dwave-qbsolv==0.3.1
dwave-system==1.2.1
dwave-tabu==0.3.0
dwavebinarycsp==0.1.2
minorminer==0.2.3
penaltymodel-cache==0.4.1
penaltymodel-lp==0.1.3
penaltymodel==0.16.3
pyqubo>=0.4.0

Extras, `all`
---

penaltymodel-mip==0.2.3
penaltymodel-maxgap==0.5.3

Changelog
---

- `pyqubo` pinned to 0.4.0

Changelog
---

* `dimod` 0.9.5 -> 0.9.7
- `Structured.to_networkx_graph` mixin method
- `.min`, `.max`, and `.sum` methods on BQM views
- `nonblocking_sample_method` decorators for asynchronous samplers
- `anti_crossing_loops`, `anti_crossing_clique`, and `doped` problem generators
- `BQM.to_file` and `BQM.from_file` methods for cyBQMs
- `VartypeView.relabel_variables` no longer sometimes returnes models with the wrong vartype
- `BQM.relabel_variables_as_integers` now works correctly when `inplace=False`
- New `DISCRETE` vartype for sample sets
- `DiscreteQuadraticModel` class
- New parallel constructor for dense c++ BQMs

* `dwave-cloud-client` 0.7.7 -> 0.8.1
- `Client` instance defaults via `defaults` constructor argument (425)
- `Client` class defaults in `Client.DEFAULTS`
- Parsing of boolean options in configuration files. `on`/`off`, `true`/`false`, `1`/`0` all work as expected (113)
- `deprecated` utility decorator
- `aliasdict` utility `dict` subclass enabling lazy access to alias keys
- Removed legacy config file format (`.dwrc`) support (138)
- Removed some of the deprecated `Solver.is_*` properties
- Removed support for deprecated formats: `bqm-ref` and `bq-zlib`.
- Removed support for more than first three positional arguments in `Client` constructor (`endpoint`, `token` and `solver`). Positional arguments in the constructor will be deprecated eventually.
- `Client.solvers` scheduled for removal in 0.9.0
- Deprecated `Future.occurrences` in favor of `Future.num_occurrences`
- Deprecated `samples` and `occurrences` keys in `Future`/`Future.result()`
- A discrete quadratic model solver, `dwave.cloud.solver.DQMSolver`
- New name for binary quadratic solver, `dwave.cloud.solver.BQMSolver`
- `Client.get_solver` has been known to mutate the `default_solver` (428). Now we make sure it doesn't.
- Setting kwargs explicitly to `None` in `Client.from_config` could overwrite config values (430). Now that too is fixed.

* `dwave-greedy` 0.1.1
- `greedy.SteepestDescentSolver`, a steepest descent greedy solver for binary quadratic models
- `greedy.SteepestDescentComposite`, a greedy descent post-processing composite

* `dwave-hybrid` 0.5.1 -> 0.6.0
- Option to retry if embedding fails in `QPUSubproblemAutoEmbeddingSampler` (114)
- `ComponentDecomposer` for iterating over problem graph components (209)
- [`greedy`](https://docs.ocean.dwavesys.com/projects/greedy/en/latest/)-based `SteepestDescentProblemSampler`/`SteepestDescentSubproblemSampler`
- `SplatComposer` was producing `numpy.float64` samplesets (235). Now it inherits `dtype` from input.

* `dwave-inspector` 0.2.3 -> 0.2.4
- Stop using hierarchical requirements include
- Support changes in latest `dwave-cloud-client` and `dwave-system`

* `dwave-neal` 0.5.4 -> 0.5.6
- Accept samples-like for `initial_states` (use `dimod.Initialized` ABC)
- Updated Windows tests to work with latest AppVeyor setup
- Stopped using hierarchical requirements include
- Skip deploying linux/py39 package, as not all dependencies are available for py39.

* `dwave-networkx` 0.8.7 -> 0.8.8
- Rearranged requirements files to follow Ocean style

* `dwave-qbsolv` 0.3.0 -> 0.3.1
- Test with dimod 0.9.x by default
- Stop using hierarchical requirements include
- Simplified and updated Windows tests to work with latest AppVeyor setup
- Skip deploying linux/py39 package, as not all dependencies are available for py39.

* `dwave-system` 0.9.9->1.1.1
- `DWaveSampler.to_networkx_graph` method
- A faster implementation of `embed_bqm`
- A `EmbeddedStructure` class.
- Improved chain strength in `DWaveCliqueSampler`
- `DWaveCliqueSampler` now correctly supports legacy software solvers
- New chain strength tuning methods, `uniform_torque_compensation` and `scaled`
- Embedding composites now use `uniform_torque_compensation` to set the default chain strength
- `MockDWaveSampler.sample` now accepts `**kwarg` for greater flexibility when testing
- All samplers now strongly prefer either a QPU or a hybrid solver. Please see [Sampler behavior changes in 1.0.0](https://github.com/dwavesystems/dwave-system/wiki/Sampler-behavior-changes-in-1.0.0) for details.
- Removed deprecated `solver_features` kwarg in `DWaveSampler`
- Removed deprecated `order_by` kwarg in `DWave(Clique)Sampler`
- Added `LeapHybridDQMSampler`

* `minorminer` 0.2.0 -> 0.2.3
- Fix for the false K_4 embeddings

Package versions
---

dimod==0.9.7
dwave-cloud-client==0.8.1
dwave-greedy==0.1.1
dwave-hybrid==0.6.0
dwave-inspector==0.2.4
dwave-neal==0.5.6
dwave-networkx==0.8.8
dwave-qbsolv==0.3.1
dwave-system==1.1.1
dwave-tabu==0.3.0
dwavebinarycsp==0.1.2
minorminer==0.2.3
penaltymodel-cache==0.4.1
penaltymodel-lp==0.1.3
penaltymodel==0.16.3
pyqubo>=0.4.0

Extras, `all`
---

penaltymodel-mip==0.2.3
penaltymodel-maxgap==0.5.3