Ankur Ankan
Ankur Ankan
`to_junction_tree` needs more tests covering the issue #889. It has been fixed in #929.
### Subject of the issue We have some explicit gradient functions for gaussian distribution https://github.com/pgmpy/pgmpy/blob/dev/pgmpy/sampling/base.py#L97. If we replace this with one of the auto grad tools (pytorch or tensorflow) it...
### Subject of the issue Reader classes should keep the original state names instead of assigning integer values to the states. ### Your environment * pgmpy version: 0.1.7 * Python...
The docstring of `get_cpds` in `BayesianModel` doesn't specify that it returns a TabularCPD object when `node` is not `None. Specify this in the docstrings and also need to improve doctests.
https://ci.appveyor.com/project/ankurankan/pgmpy/build/1.0.4 ``` ====================================================================== FAIL: test_sampling (pgmpy.tests.test_sampling.test_continuous_sampling.TestNUTSInference) ---------------------------------------------------------------------- Traceback (most recent call last): File "C:\projects\pgmpy-pgmh6\pgmpy\tests\test_sampling\test_continuous_sampling.py", line 139, in test_sampling self.assertTrue(np.linalg.norm(sample_covariance - self.test_model.covariance) < 0.4) AssertionError: False is not true ---------------------------------------------------------------------- ```...
https://github.com/pgmpy/pgmpy_notebook/pull/7
Recently we refactored pgmpy (https://github.com/pgmpy/pgmpy/pull/731) so we need to update the notebooks according to that.
```python In [1]: from pgmpy.global_vars import config In [2]: config.set_backend('torch') In [3]: config.get_backend() Out[3]: 'torch' In [4]: from pgmpy.utils import get_example_model In [5]: model = get_example_model('alarm') df In [6]: df...
### Your checklist for this pull request Please review the [guidelines for contributing](CONTRIBUTING.md) to this repository. - [ ] Make sure you are requesting to **pull a topic/feature/bugfix branch** (right...
```python In [1]: from pgmpy.models import BayesianNetwork In [2]: model = BayesianNetwork([('A', 'B'), ('A', 'C'), ('B', 'D'), ('C', 'D')]) In [3]: from pgmpy.factors.discrete import TabularCPD In [4]: model = model.get_random_cpds()...