4e6527c | Thomas Wiecki | 03 June 2022, 15:51:31 UTC | Bump version to 4.0.0 🎉 | 03 June 2022, 15:51:31 UTC |
e663f46 | Thomas Wiecki | 03 June 2022, 15:23:57 UTC | Remove developer guide. Needs to be updated eventually. | 03 June 2022, 15:50:44 UTC |
938604c | Ricardo | 03 June 2022, 14:16:35 UTC | Add temporary fix for pretty representation of SymbolicDistributions | 03 June 2022, 15:13:01 UTC |
57c6a8f | Ricardo | 03 June 2022, 13:55:41 UTC | Move printing-related tests to their own test module | 03 June 2022, 15:13:01 UTC |
c758d0d | Alexandre Andorra | 03 June 2022, 14:43:38 UTC | Restructure Release Notes (#5832) * 1st try at restructuring * Add emojis and do-not-miss section * Apply suggestions from code review Co-authored-by: Michael Osthege <michael.osthege@outlook.com> Co-authored-by: Ricardo Vieira <28983449+ricardoV94@users.noreply.github.com> * Improve structure based on review * Apply suggestions from code review Co-authored-by: Ricardo Vieira <28983449+ricardoV94@users.noreply.github.com> * Improve structure again * Add release announcement * Apply suggestions from code review Co-authored-by: Thomas Wiecki <thomas.wiecki@gmail.com> * Apply suggestions from code review Co-authored-by: Thomas Wiecki <thomas.wiecki@gmail.com> * Apply suggestions from code review Co-authored-by: Thomas Wiecki <thomas.wiecki@gmail.com> * Add highlight of JAX backend. * Remove BART section and move links to where we remove it. * Move BART further down. Remove mention of size. Co-authored-by: Michael Osthege <michael.osthege@outlook.com> Co-authored-by: Ricardo Vieira <28983449+ricardoV94@users.noreply.github.com> Co-authored-by: Thomas Wiecki <thomas.wiecki@gmail.com> | 03 June 2022, 14:43:38 UTC |
0d38abf | Ravin Kumar | 03 June 2022, 14:08:32 UTC | Move BMCP to top of book list | 03 June 2022, 14:28:34 UTC |
5da32ed | Ricardo | 03 June 2022, 12:27:42 UTC | Add comment to `set_data` condition where the length comes from a `shape[i]` graph | 03 June 2022, 14:28:09 UTC |
70b0c49 | Michael Osthege | 30 May 2022, 18:44:41 UTC | Automatically resize independently mutable dims via `set_data` | 03 June 2022, 14:28:09 UTC |
2120593 | Michael Osthege | 28 May 2022, 16:33:12 UTC | Warn about resizing `MutableData` dims that are not symbolically linked Closes #5812 | 03 June 2022, 14:28:09 UTC |
a74592a | Michael Osthege | 28 May 2022, 16:32:15 UTC | Add `Model.set_dim` method for safer dims resizing | 03 June 2022, 14:28:09 UTC |
0e7cd1f | Michael Osthege | 26 May 2022, 20:00:27 UTC | Update guide on implementing new distributions Co-authored-by: Oriol Abril-Pla <oriol.abril.pla@gmail.com> | 03 June 2022, 13:38:09 UTC |
2dc0009 | Michael Osthege | 26 May 2022, 19:23:06 UTC | Add troubleshooting tips on pre-commit | 03 June 2022, 13:38:09 UTC |
48eff60 | Michael Osthege | 26 May 2022, 18:58:39 UTC | Write instructions on running the test suite | 03 June 2022, 13:38:09 UTC |
559969c | Ricardo | 10 May 2022, 16:14:40 UTC | Update Aesara dependency | 02 June 2022, 22:37:35 UTC |
2d6304d | Oriol (ZBook) | 02 June 2022, 17:54:51 UTC | fix 404 and intersphinx | 02 June 2022, 22:33:32 UTC |
88a0f43 | Thomas Wiecki | 02 June 2022, 19:39:29 UTC | Bring back learn page. (#5833) * Bring back learning page. * Remove no-sidebar. * Updates to user guide. * Remove learn.md. * Add link to PyMC Labs. * Link to correct image. * Improve cosulting page. Rename to learn. * Change cards to use Oriol's proposed format. * fix sphinx build * add newline at end of file * ensure img kwarg is in the next line of directive definition Co-authored-by: Oriol Abril-Pla <oriol.abril.pla@gmail.com> | 02 June 2022, 19:39:29 UTC |
de9db67 | Purna Chandra Mansingh | 02 June 2022, 16:05:04 UTC | Deprecate remaining VI methods (#5806) * Deprecated remaining VI methods * Deleted flows and code for local and batched kwarg * removed rowwise from opvi.py * redo the test_variational_inference.py * refactored some test cases * removed mention of local from docstring * refactor some test in varialional_inference * did correction in docstring * removed batched dimension property * added symbolic_random * removed part = 'total' * refactor test_logq_globals and test_pickle_approx test * refactor test_init_groups | 02 June 2022, 16:05:04 UTC |
b073fd8 | Ricardo | 02 June 2022, 09:05:28 UTC | Remove user-facing mentions of size in timeseries distributions | 02 June 2022, 16:02:50 UTC |
343a6f5 | Ricardo | 31 May 2022, 15:15:28 UTC | Use more wide default init_dist in GaussianRandomWalk and AR and raise UserWarning when not explicitly defined | 02 June 2022, 16:02:50 UTC |
12039e0 | Ricardo | 31 May 2022, 15:14:38 UTC | Rename GaussianRandomWalk `init` argument to `init_dist` and fix some type hints | 02 June 2022, 16:02:50 UTC |
42e3745 | Ricardo | 02 June 2022, 08:51:34 UTC | Do not issue warning when default transform of Mixture is None Also fixes typo in warning message | 02 June 2022, 14:22:05 UTC |
db5c87e | Oriol Abril-Pla | 01 June 2022, 10:38:10 UTC | Update navbar (#5830) * start updating navbar * get navbar and redirects to work * make search bar return the same results from all websites * update requirements * remove search bar from sidebar * fix urls and add extra page * Update .pre-commit-config.yaml Co-authored-by: Thomas Wiecki <thomas.wiecki@gmail.com> | 01 June 2022, 10:38:10 UTC |
1e7d91f | Ricardo | 30 May 2022, 07:48:13 UTC | Make Metropolis cope better with multiple dimensions Metropolis now updates each dimension sequentially and tunes a proposal scale parameter per dimension | 31 May 2022, 21:20:17 UTC |
4810829 | Ricardo | 31 May 2022, 07:21:39 UTC | Move TestMetropolisProposal tests to TestMetropolis | 31 May 2022, 21:20:17 UTC |
6750ea7 | Osvaldo A Martin | 31 May 2022, 17:23:07 UTC | Supress warning SMC initialization (#5827) | 31 May 2022, 17:23:07 UTC |
6027be1 | Ricardo | 31 May 2022, 14:09:37 UTC | Raise NotImplementedError in sample_posterior_predictive_w | 31 May 2022, 16:26:29 UTC |
ae578c2 | Maxim Kochurov | 24 May 2022, 11:23:57 UTC | Raise NotImplementedError in not yet refactored timeseries distributions Co-authored-by: Thomas Wiecki <thomas.wiecki@gmail.com> | 31 May 2022, 16:26:29 UTC |
5703a9d | Alexandre Andorra | 30 May 2022, 18:03:08 UTC | Remove advanced Aesara usage guide (#5826) | 30 May 2022, 18:03:08 UTC |
6b22ed5 | Michael Osthege | 30 May 2022, 17:44:55 UTC | Add return type hints (#5819) * Add return type hints * Add type hints for rv arguments * Fix more type issues Closes #4880 Co-authored-by: Ricardo Vieira <28983449+ricardoV94@users.noreply.github.com> | 30 May 2022, 17:44:55 UTC |
57654dc | Ricardo | 20 January 2022, 18:25:14 UTC | Add name to CompountStep | 30 May 2022, 16:28:55 UTC |
35393af | Ricardo | 28 May 2022, 16:19:31 UTC | Avoid unnecessary slow compilation of Aesara functions in test_step | 30 May 2022, 16:28:55 UTC |
e636852 | Ricardo | 20 January 2022, 18:24:47 UTC | Refactor TestStepMethods and reduce number of draws The number of draws was set too high to accommodate the worst / buggy samplers (see #5815) | 30 May 2022, 16:28:55 UTC |
70a3e0b | Ricardo | 28 May 2022, 16:00:36 UTC | Refactor Slice.astep method Avoid recreation of RaveledArrays in inner loops and repeated indexing of `self.w` | 30 May 2022, 16:28:55 UTC |
f28f9cf | Ricardo | 28 May 2022, 16:10:00 UTC | Fix Slice sampler bugs introduced in b988ba99a084a18b2c907ef2ee4aeb1f4ab56160 1. The previous interval was of the form `[q0 - uniform(0, self.w), q0 + self.w]` which is not symmetric around q0 in expectation, breaking detailed balance. 2. The left/right bound values used for updating earlier variables were only set to the accepted values during tuning. This might have made bug 1 worse. | 30 May 2022, 16:28:55 UTC |
de55543 | Danh Phan | 30 May 2022, 09:54:42 UTC | Add Quick links to Contributing guidelines (#5818) | 30 May 2022, 09:54:42 UTC |
4d2f3a8 | Ricardo | 25 May 2022, 15:38:01 UTC | Fallback to Aeppl when logp is not implemented | 26 May 2022, 12:24:49 UTC |
7d7d07c | Ricardo | 26 May 2022, 07:15:16 UTC | Update Aeppl dependency | 26 May 2022, 12:24:49 UTC |
7b239bb | Michael Osthege | 25 May 2022, 16:38:58 UTC | Limit model graph nodes to named model variables Some nodes such as `NoneConst` in an Aesara graph are named, but aren't model variables that the user is interested in. | 26 May 2022, 11:06:24 UTC |
2161664 | Michael Osthege | 25 May 2022, 15:44:46 UTC | Forward `dims` to reparametrized `Deterministic` This way `.prior(..., dims=...)` applies not only to the `_rotated_` RV but also to the `Deterministic` that the user is interested in. | 26 May 2022, 10:26:41 UTC |
9ca57bf | Michael Osthege | 25 May 2022, 15:40:51 UTC | Fix shape problem in GP priors Shape asserts were added in existing tests that covered both reparameterize=True/False. Closes #5803 | 26 May 2022, 10:26:41 UTC |
b67a76f | Michael Osthege | 26 May 2022, 08:57:42 UTC | Pin blackjax version for testing to fix failing CI Related to https://github.com/blackjax-devs/blackjax/issues/212 | 26 May 2022, 10:26:41 UTC |
2adb346 | Ricardo Vieira | 24 May 2022, 20:45:35 UTC | Fix ReleaseNotes | 25 May 2022, 04:41:03 UTC |
62e1e9d | Ricardo | 23 May 2022, 11:05:09 UTC | Seed sampling in TestNutsCheckTrace.test_linalg | 24 May 2022, 20:33:39 UTC |
15431d9 | Ricardo | 23 May 2022, 12:08:22 UTC | Fix non-deterministic NUTS initialization when using ADVI | 24 May 2022, 20:33:39 UTC |
cee2b30 | Ricardo | 23 May 2022, 13:04:57 UTC | Remove model.rng_seeder and allow compile_pymc to reseed RNG variables in compiled function Sampling functions now also accept RandomState or Generators as input to random_seed, similarly to how random_state behaves in scipy distributions. For backwards compatibility this argument was not renamed. | 24 May 2022, 20:33:39 UTC |
09240fc | Ricardo | 23 May 2022, 10:14:44 UTC | Update expected behavior of `sample`-related tests in relation to global seeding. Together, `test_sample_does_not_set_seed` and `test_parallel_sample_does_not_reuse_seed` covered two unspoken behaviors of `sample`: 1. When no seed is specified, PyMC shall not set global seed state of numpy in the main process. 2. When no seed is specified, sampling will depend on numpy global seeding state for reproducible behavior. Point 1 is due to PyMC legacy dependency on global seeding for step samplers. It tries to minimize "damage" by only setting global seeds when it absolutely needs to, in order to ensure deterministic sampling. Ideally calls to `numpy.seed` would never be made. Point 2 goes against NumPy current best practices of using None when defining new Generators / SeedSequences (https://numpy.org/doc/stable/reference/random/bit_generators/generated/numpy.random.SeedSequence.html#numpy.random.SeedSequence) The refactored tests cover point 1 more directly, and assert the opposite of point 2. | 24 May 2022, 20:33:39 UTC |
c71fd4e | Ricardo | 23 May 2022, 06:56:38 UTC | Fix default use of global seeding | 24 May 2022, 20:33:39 UTC |
4caa5b9 | Ricardo | 19 May 2022, 15:36:10 UTC | Refactor RandomState/Generator reseeding logic in initial_point into aesaraf | 24 May 2022, 20:33:39 UTC |
b5a5b56 | Michael Osthege | 23 May 2022, 13:49:41 UTC | Remove size from Distribution signatures (#5788) * Remove size from .dist() signature Closes #5754 * Align (Half)Flat signatures with superclass * Don't mention size in the docstring Co-authored-by: Ricardo Vieira <28983449+ricardoV94@users.noreply.github.com> * Revert changes in `rng_fn` signatures Co-authored-by: Ricardo Vieira <28983449+ricardoV94@users.noreply.github.com> Co-authored-by: Ricardo Vieira <28983449+ricardoV94@users.noreply.github.com> | 23 May 2022, 13:49:41 UTC |
d0af6b1 | larryshamalama | 22 May 2022, 11:23:21 UTC | New `var_names` kwarg for `pm.model_to_graphviz` (#5634) Enables positive selection of model variables to be included in the model graph. Co-authored-by: Michael Osthege <michael.osthege@outlook.com> | 22 May 2022, 11:23:21 UTC |
66fba38 | Ricardo Vieira | 22 May 2022, 07:47:33 UTC | Add warning about unobserved continuous censored distributions Co-authored-by: Michael Osthege <michael.osthege@outlook.com> Co-authored-by: Oriol Abril-Pla <oriol.abril.pla@gmail.com> Co-authored-by: Michael Osthege <michael.osthege@outlook.com> | 22 May 2022, 07:47:33 UTC |
a95dd71 | Luke LB | 14 May 2022, 14:16:44 UTC | Require `pm.MutableData` or `add_coord(mutable=True)` for mutable dims This increases the safety of working with resizable dimensions. Dims created via `pm.Model(coords=...)` coordinates are now immutable. Only dimension lengths that are created anew from `pm.MutableData`, `pm.Data(mutable=True)`, or the underlying `add_coord(mutable=True)` will become shared variables. Co-authored-by: Michael Osthege <michael.osthege@outlook.com> | 21 May 2022, 19:47:02 UTC |
7ec106c | Michael Osthege | 21 May 2022, 10:54:10 UTC | Optionally pass `coords` when creating `pm.Data` variables Needed to create resizable dims with coordinate values, because coords passed to the model will become immutable dims. See https://github.com/pymc-devs/pymc/pull/5763. Also, because `pm.Data` variables can be N-dimensional, the coordinate values can't reliably be taken from the value. | 21 May 2022, 19:47:02 UTC |
e83520e | Luke LB | 16 May 2022, 20:51:33 UTC | Remove ShapeWarning in favor of raising ValueError directly | 21 May 2022, 19:47:02 UTC |
2b7a3c2 | Oriol Abril-Pla | 20 May 2022, 15:16:31 UTC | add calendar to website (#5786) * add calendar to website * update navbar about item to pymc.io * update urls in intersphinx mapping * add description above the calendar | 20 May 2022, 15:16:31 UTC |
b91283e | Ricardo | 20 January 2022, 18:28:45 UTC | Pass seed to `find_map` and set global seeds in `_iter_sample` and `_prepare_iter_sample` This reverts some changes in 47b61def5f24830e36714886356557a0e3918195 which wrongly disabled global seeding in some sampling contexts that still depended on it. | 19 May 2022, 22:05:24 UTC |
56ad6a9 | Ricardo | 20 January 2022, 18:28:18 UTC | Add explanation to test_sample_does_not_set_seed | 19 May 2022, 22:05:24 UTC |
f0c8f55 | Juan Orduz | 19 May 2022, 22:03:57 UTC | Update pr_tutorial.md (#5724) * Update pr_tutorial.md Make the python version consistent on the PR tutorial. * update to python 3.9 | 19 May 2022, 22:03:57 UTC |
00d4372 | Ricardo | 18 May 2022, 08:27:13 UTC | Update distribution-related changes in RELEASE-NOTES.md | 19 May 2022, 05:48:12 UTC |
c2d96cf | Ricardo | 18 May 2022, 07:36:41 UTC | Use unit normal as default init_dist in GaussianRandomWalk and AR | 19 May 2022, 05:48:12 UTC |
2536920 | Oriol Abril-Pla | 19 May 2022, 04:03:35 UTC | add subsections inside community section (#5782) | 19 May 2022, 04:03:35 UTC |
82f70fc | Ricardo | 18 May 2022, 08:01:11 UTC | Filter deterministics from initial_points returned by init_nuts | 18 May 2022, 09:35:43 UTC |
817f457 | Ricardo | 17 May 2022, 12:24:20 UTC | Remove size parameter from sample_posterior_predictive | 18 May 2022, 08:05:20 UTC |
862bd05 | lucianopaz | 12 May 2022, 14:16:09 UTC | Add compile_forward_sampling_function | 17 May 2022, 16:55:36 UTC |
4969460 | Ricardo | 17 May 2022, 12:39:55 UTC | Move Identity operator initialization close to class definition | 17 May 2022, 14:40:33 UTC |
c327a02 | Ricardo | 17 May 2022, 12:38:08 UTC | Remove take_along_axis in favor of Aesara's implementation | 17 May 2022, 14:40:33 UTC |
8520a2c | Ricardo | 10 May 2022, 16:47:49 UTC | Make test_sample_posterior_predictive_after_set_data_with_coords more robust Closes #5739 | 17 May 2022, 13:13:39 UTC |
85468b0 | TimOliverMaier | 16 May 2022, 17:41:48 UTC | Improve errors and warnings related to dimension resizing (#5761) * Support TensorConstant entries in `Model.dim_lengths`. * Raise errors on attempts to resize dims with `TensorConstant` length. * Raise errors on attempts to resize dims that are linked to non-shared variables. * Warn about resizing dims that weren't initialized from a shared variable (supposedly via `add_coord(..., fixed=False)`, see #5763). * Raise errors when attempting to resize a dim that had coord values without providing new coord values. Closes #5760 by anticipating that not all symbolic dim lengths originate from RVs. Co-authored-by: Michael Osthege <michael.osthege@outlook.com> | 16 May 2022, 17:41:48 UTC |
3b4da05 | Josh Cook | 15 May 2022, 04:14:38 UTC | doc: add `nuts_kwargs` to `sample_numpyro_nuts` docstring (#5764) The `nuts_kwargs` argument is currently missing from the docstring of `sample_numpyro_nuts`. This just adds a simple statement to indicate that the keyword arguments are passed to `numpyro.infer.NUTS()`. | 15 May 2022, 04:14:38 UTC |
ab593b1 | Michael Osthege | 09 May 2022, 18:57:20 UTC | Remove `EllipticalSlice` sampler Closes #5137 | 09 May 2022, 19:58:08 UTC |
ce2e891 | Christian Luhmann | 09 May 2022, 11:23:41 UTC | Update doc links in README (#5753) * Update links to overview and examples. * Overview links fixed. | 09 May 2022, 11:23:41 UTC |
e7bfc25 | twiecki | 09 May 2022, 07:04:08 UTC | ⬆️ UPGRADE: Autoupdate pre-commit config | 09 May 2022, 11:23:06 UTC |
49eecfb | Christian Ritter | 08 May 2022, 12:42:12 UTC | Introduced tests for pymc.distributions.logprob._get_scaling_ (#5544) * Introduced test for _get_scaling_ Co-authored-by: Icyshaman <vikasvishwakarma6787@gmail.com> Co-authored-by: chritter <christian.ritter@canada.ca> | 08 May 2022, 12:42:12 UTC |
53ad689 | Chris Fonnesbeck | 06 May 2022, 17:30:44 UTC | Refactor get_tau_sigma to handle lists (#5747) * Refactor get_tau_sigma to handle lists * Fixed syntax error * Black formatted * Change assertions to ValueError raises * Prior predictions constant data (#5723) * support return constant_data with prior pred * pre-commit format * added test * code review * Fix issue probably-meant-fstring found at https://codereview.doctor (#5726) * Add coords argument to pymc.set_data (#5588) Co-authored-by: Oriol (ZBook) <oriol.abril.pla@gmail.com> * remove MultinomialRV override * ⬆️ UPGRADE: Autoupdate pre-commit config * Group GaussianRandomWalk tests in single class * Infer steps from shape in GaussianRandomWalk * Unpin upper limit on numpydoc Closes #5401 * ⬆️ UPGRADE: Autoupdate pre-commit config * Standardize docstrings of input dists arguments and add warning about cloning * Remove unnecessary tag in Simulator logp * Replace deprecated tag.ignore_logprob * Group compile_pymc tests in own class * Remove remaining uses of default_updates in codebase * Remove redundant/wrong docstrings from GaussianRandomWalk logp * Add moment to GaussianRandomWalk and fix mu/sigma broadcasting bug * Do not create temporary SymbolicDistribution just to retrieve number of RNGs needed Reordered methods for consistency * Move SymbolicDistribution docstring to body of class * Refactor AR distribution * Deprecates AR1 distribution * Implements random and moment methods * Batching works on the left as with other distributions * Rename `pandas_to_array` to `convert_observed_data` * Obtain step information from dims and observed * Make AR steps extend shape beyond initial_dist This is consistent with the meaning of steps in the GaussianRandomWalk and translates directly to the number of scan steps taken Co-authored-by: TimOliverMaier <tmaier01@students.uni-mainz.de> Co-authored-by: code-review-doctor <72647856+code-review-doctor@users.noreply.github.com> Co-authored-by: Somasree Majumder <56045049+soma2000-lang@users.noreply.github.com> Co-authored-by: Oriol (ZBook) <oriol.abril.pla@gmail.com> Co-authored-by: danhphan <danh.phan.mq@gmail.com> Co-authored-by: twiecki <twiecki@users.noreply.github.com> Co-authored-by: Ricardo <ricardo.vieira1994@gmail.com> Co-authored-by: Ravin Kumar <ravinsdrive@gmail.com> Co-authored-by: Michael Osthege <michael.osthege@outlook.com> | 06 May 2022, 17:30:44 UTC |
dd23c90 | Ricardo | 05 May 2022, 13:09:06 UTC | Make AR steps extend shape beyond initial_dist This is consistent with the meaning of steps in the GaussianRandomWalk and translates directly to the number of scan steps taken | 06 May 2022, 11:37:19 UTC |
812a52c | Ricardo | 05 May 2022, 09:31:54 UTC | Obtain step information from dims and observed | 06 May 2022, 11:37:19 UTC |
9a905fa | Ricardo | 05 May 2022, 09:22:28 UTC | Rename `pandas_to_array` to `convert_observed_data` | 06 May 2022, 11:37:19 UTC |
c76b9b9 | Ricardo | 29 April 2022, 15:04:55 UTC | Refactor AR distribution * Deprecates AR1 distribution * Implements random and moment methods * Batching works on the left as with other distributions | 05 May 2022, 08:02:42 UTC |
c18e6ce | Ricardo | 04 May 2022, 15:21:58 UTC | Move SymbolicDistribution docstring to body of class | 05 May 2022, 08:02:42 UTC |
4471204 | Ricardo | 29 April 2022, 15:10:07 UTC | Do not create temporary SymbolicDistribution just to retrieve number of RNGs needed Reordered methods for consistency | 05 May 2022, 08:02:42 UTC |
21c2e6c | Ricardo | 04 May 2022, 10:41:28 UTC | Add moment to GaussianRandomWalk and fix mu/sigma broadcasting bug | 04 May 2022, 14:49:45 UTC |
1a5b3d6 | Ricardo | 04 May 2022, 10:05:56 UTC | Remove redundant/wrong docstrings from GaussianRandomWalk logp | 04 May 2022, 14:49:45 UTC |
a3bd083 | Ricardo | 03 May 2022, 10:02:10 UTC | Remove remaining uses of default_updates in codebase | 04 May 2022, 11:54:11 UTC |
ee73231 | Ricardo | 03 May 2022, 09:42:17 UTC | Group compile_pymc tests in own class | 04 May 2022, 11:54:11 UTC |
138a64f | Ricardo | 03 May 2022, 09:09:42 UTC | Replace deprecated tag.ignore_logprob | 04 May 2022, 11:54:11 UTC |
b05af43 | Ricardo | 03 May 2022, 09:42:07 UTC | Remove unnecessary tag in Simulator logp | 04 May 2022, 11:54:11 UTC |
f95e839 | Ricardo | 03 May 2022, 09:32:37 UTC | Standardize docstrings of input dists arguments and add warning about cloning | 04 May 2022, 11:54:11 UTC |
4ae8b07 | twiecki | 02 May 2022, 07:04:41 UTC | ⬆️ UPGRADE: Autoupdate pre-commit config | 03 May 2022, 22:13:58 UTC |
5c39f7e | Michael Osthege | 01 May 2022, 07:43:49 UTC | Unpin upper limit on numpydoc Closes #5401 | 01 May 2022, 09:33:45 UTC |
6ae7628 | Ravin Kumar | 05 April 2022, 02:48:37 UTC | Infer steps from shape in GaussianRandomWalk | 29 April 2022, 15:30:08 UTC |
c80657e | Ricardo | 27 April 2022, 16:29:34 UTC | Group GaussianRandomWalk tests in single class | 29 April 2022, 15:30:08 UTC |
54425cd | twiecki | 25 April 2022, 07:05:52 UTC | ⬆️ UPGRADE: Autoupdate pre-commit config | 27 April 2022, 16:44:31 UTC |
458f28e | danhphan | 19 March 2022, 10:19:11 UTC | remove MultinomialRV override | 27 April 2022, 12:52:44 UTC |
bae5087 | Somasree Majumder | 24 April 2022, 08:46:27 UTC | Add coords argument to pymc.set_data (#5588) Co-authored-by: Oriol (ZBook) <oriol.abril.pla@gmail.com> | 24 April 2022, 08:46:27 UTC |
e4ec363 | code-review-doctor | 24 April 2022, 02:13:20 UTC | Fix issue probably-meant-fstring found at https://codereview.doctor (#5726) | 24 April 2022, 02:13:20 UTC |
162ad6c | TimOliverMaier | 22 April 2022, 21:16:03 UTC | Prior predictions constant data (#5723) * support return constant_data with prior pred * pre-commit format * added test * code review | 22 April 2022, 21:16:03 UTC |
d322f75 | Michael Osthege | 30 March 2022, 09:28:38 UTC | Upgrade Aesara and Aeppl dependencies | 20 April 2022, 11:48:16 UTC |
bc0ec61 | Ricardo | 20 April 2022, 10:44:59 UTC | Remove unnecessary at.squeeze in logp and logcdf of Mixtures | 20 April 2022, 11:48:16 UTC |
d60b4cf | Ricardo | 20 April 2022, 10:55:20 UTC | Correct xfail reason in test_nested_mixture | 20 April 2022, 11:48:16 UTC |
d02707e | Ricardo | 19 April 2022, 09:55:48 UTC | Move GaussianRandomWalk init resizing to Op.make_node | 19 April 2022, 19:33:23 UTC |