Hmmlearn

~~~~~~~~~~~~~

Released on October 31, 2024.

- Provide wheels compatible with numpy 2.

-------------

Released on March 1st, 2024.

- update CI/CD Pipelines that were troublesome

-------------

Released on March 1st, 2024.

- Support Python 3.8-3.12
- Improve stability of test suite. Ensure the documentation examples are covered.
- Documentation Improvements throughout.

-------------

Released on April 18th, 2023.

- Introduce learning HMMs with Variational Inference. Support
Gaussian and Categorical Emissions. This feature is provisional and subject
to further changes.
- Deprecated support for inputs of shape other than ``(n_samples, 1)`` for
categorical HMMs.
- Removed the deprecated ``iter_from_X_lengths`` and ``log_mask_zero``;
``lengths`` arrays that do not sum up to the entire array length are no
longer supported.
- Support variable ``n_trials`` in ``MultinomialHMM``, except for sampling.

-------------

Released on September 26th, 2022.

- The ``PoissonHMM`` class was added with an example use case.
- For ``MultinomialHMM``, parameters after ``transmat_prior`` are now
keyword-only.
- ``startprob_`` and ``transmat_`` will both be initialized with random
variables drawn from a Dirichlet distribution, to maintain the old
behavior, these must be initialized as ``1 / n_components``.
- The old ``MultinomialHMM`` class was renamed to ``CategoricalHMM`` (as that's
what it actually implements), and a new ``MultinomialHMM`` class was
introduced (with a warning) that actually implements a multinomial
distribution.
- ``hmmlearn.utils.log_mask_zero`` has been deprecated.

-------------

Released on February 10th, 2022.

- Dropped support for Py3.5 (due to the absence of manylinux wheel supporting
both Py3.5 and Py3.10).
- ``_BaseHMM`` has been promoted to public API and has been renamed to
``BaseHMM``.
- MultinomialHMM no longer overwrites preset ``n_features``.
- An implementation of the Forward-Backward algorithm based upon scaling
is available by specifying ``implementation="scaling"`` when instantiating
HMMs. In general, the scaling algorithm is more efficient than an
implementation based upon logarithms. See `scripts/benchmark.py` for
a comparison of the performance of the two implementations.
- The *logprob* parameter to `.ConvergenceMonitor.report` has been renamed to
*log_prob*.