eb8680
eb8680
~~Blocked for now by #51~~ We may also need a `TruncatedNormal` distribution. This issue proposes to add versions of the examples in section 6 from [Discrete-Continuous Mixtures in Probabilistic Programming:...
@jacobrgardner pointed out today that GPyTorch implements heavily optimized, numerically stable versions of the abstract Gaussian operations added in #37. We should look into using those to implement the `funsor.Gaussian`...
In #161 and #157 we've seen test failures caused by a failure to reduce expressions like `0 + Gaussian(...)` or `1 * Gaussian(...)`. We should add any missing patterns for...
There's a prototype pair-coded by @fritzo and @eb8680 in [this branch](https://github.com/pyro-ppl/funsor/tree/fixpoints), but it needs to be updated and documented. We might also want a [polyvariadic version](http://okmij.org/ftp/Computation/fixed-point-combinators.html#Poly-variadic) to express mutual recursion.
From #3192, the following `torch.jit`-related tests are failing for `torch>=2.0.0`: - [ ] `tests/contrib/test_minipyro.py::test_elbo_jit[minipyro]` ([traceback](https://github.com/pyro-ppl/pyro/actions/runs/5008810154/jobs/8977127685?pr=3192#step:5:13814)) - [ ] `tests/contrib/cevae/test_cevae.py::test_serialization[jit-exponential-1]` ([traceback](https://github.com/pyro-ppl/pyro/actions/runs/5008810154/jobs/8977127685?pr=3192#step:5:13818)) - [ ] `tests/contrib/cevae/test_cevae.py::test_serialization[jit-exponential-2]` ([traceback](https://github.com/pyro-ppl/pyro/actions/runs/5008810154/jobs/8977127685?pr=3192#step:5:13882)) - [ ] `tests/contrib/test_minipyro.py::test_elbo_equivalence` The...
From #3192, the following distribution tests fail under `torch>=2.0.0` and should be fixed prior to the next release: - [ ] `tests/distributions/test_rejector.py::test_rejector` - [ ] `tests/distributions/test_stable.py::test_additive` - [ ] `tests/infer/reparam/test_stable.py::test_stable`...
brms supports autoregressive correlation structure on response variables via the `cor` argument to its `fit` function and the [`cor_arma` object](https://paul-buerkner.github.io/brms/reference/cor_ar.html). It would be nice to support at least a subset...
The [`mvbrmsformula` function](https://paul-buerkner.github.io/brms/reference/mvbrmsformula.html) in `brms` provides [support for multiple formulas with shared inputs](https://paul-buerkner.github.io/brms/articles/brms_multivariate.html): ``` xf
Currently brmp expects data in the form of Pandas dataframes, but when working with larger, higher-dimensional datasets it might be more convenient to specify formulas in terms of multidimensional [xarray](http://xarray.pydata.org/en/stable/index.html)...
Extends #43. In addition to supporting brms' existing model-based imputation feature #43 , I would like to propose adding support for multiple responses #42 where a strict subset of responses...