-
Hi,
there is another conceptual thing that we came across when comparing likelihood values between the frameworks. It concerns log(or log10)-transformed observables. In this case we have a sigma fo…
-
As Kerby pointed out on the mailing list, once we have constrained estimation, we can invert the LR test to obtain confidence intervals for individual parameters
for example for GLM
http://support.sa…
-
I'm trying to think of a clean solution for this one that isn't just "make the user write the distribution sampling function themselves".
Are custom likelihoods a niche enough feature that this lev…
-
# Context
Documenting field-level explicit likelihood inference from a differentiable cosmological model.
In [code](https://github.com/hsimonfroy/montecosmo/blob/a7346788b5555b2f6b14bff3e9dc2f4c9…
-
Right now, all of our explorers fail if the initialization state has infinite target logpotential. When this is not due to coding issues, the most likely cause is that the support of the likelihood is…
-
Would be more clear if all parameters were defined at the top of a setup file, as opposed to at several places as the likelihoods are being setup. See comments below.
```
data = pd.read_csv(os.pat…
-
# Documentation/tutorial notebooks
As of `2.9.1`, the `log_marginal_likelihood` is deprecated.
See the docs [here](https://gpflow.github.io/GPflow/2.9.1/api/gpflow/models/gpr/index.html#gpflow.mo…
-
-
Add the ROQ likelihood
-
Hey there
I wanted to ask whether it's feasible to optimise a nondifferentiable posterior. For large macro models it's often impossible to derive that gradient, either manually or via AD. Like if th…