Jan
Jan
more sources for inspiration: - https://estimagic.org/en/latest/ - https://pydvl.org/stable/
Hi @evavanweenen thanks a lot for this feedback! This is very helpful. We will have a look at this at the hackathon in March.
replaced by #1341
In the refactoring #1196 we already separated the sampling from the `run_{sbc, tarp}`. So in principle, it's possible to do things more low-level, e.g., first generate posterior samples on a...
> Update after talking to @manuelgloeckler, the easiest way to calculate the MAP here would be to use the score directly at a time t = epsilon, instead of calculating...
Yes, MNLE implements a mixed estimator with a categorical distribution. It would be nice to extend this to discrete flows as well.
The MNLE classes are now refactored to match the API of the other build functions, including z-scoring and embedding nets. Thus, in principle, one can now also use the MNLE...
We now have discrete estimators for multiple dimensions, see #1269 and this issue will be fixed by #1362
Note that `nflows` (and `sbi` posteriors) enable passing a batch of conditions to `sample`. The return shape is ```python (num_conditions, num_samples, x_shape) = flow.sample(num_samples, condition) ``` with `condition.shape=(num_conditions, x_shape)`.
will be fixed with #1153