beanmachine
beanmachine copied to clipboard
A library that allows for inference on probabilistic models
### Issue Description Need to pin pytorch to 1.13.1 so tests pass. ### Steps to Reproduce ### Expected Behavior ### System Info Please provide information about your setup - PyTorch...
### Motivation This removes arviz as an explicit dependency of Bean Machine. This will also prevent a circular dependency with arviz. ### Changes proposed Updates setup.py and moves import into...
### Motivation In preparation for a refactoring of the Bean Machine test suite, #1674 and #1711 moved the Pytest modules (BM/Python) into a dedicated `tests/` directory. This is now followed...
Summary: Adds a new transform, Sigma, which maps a probability value in the range (0..1) to the range (-INF..INF). This is one of the blockers preventing distributions like Beta from...
Summary: Implement the scalar operator LOG1P for log(1 + x) And the corresponding elementwise matrix operator MATRIX_LOG1P Reviewed By: ericlippert Differential Revision: D38923743
Summary: The compiler should detect when a model uses an operation which we know for certain cannot be accumulated into the graph and immediately throw a helpful error message. This...
Summary: We represent edge requirements imposed by BMG by a requirement object; there is no reason why we ever need two or more instances of the object representing "there is...
### Motivation This uses scikit-learn to show what happens when a regular regression is used and motivates the need for robustness in some examples with a plot. ### Changes proposed...
Summary: We're getting to the end of this series of refactorings. In this diff I start adding *combinators* on *graph fixers*. Recall that a graph fixer is a pass which...
Summary: In this diff I complete the initial round of refactorings of the graph fixing passes, by refactoring the passes which turn an observe-true on a Bernoulli into a factor...