Christoph Weniger
Christoph Weniger
Goal: Strategy for generating posteriors for parameters with implicit priors. - Parameters are part of the model prediction - Original priors and simulations are defined as before - Dataset is...
We should explore if replacing some of the network structures with Lazy initialization could simplify writing head-networks etc. We might get around having to specify `self.n_features` in that way.
Currently Swyft crashes with PL 2.0.0. Swyft should be updated without breaking backward compatibility with PL 1.9.0 and earlier.
In order to make that happen, we need to consolidate the Gaussian_Autoregressive interface. What about dropping the `_X`, and only keeping the module-based version? We can later add functionality to...
Right now we have plot2.py, plot.py, incoherent usage of labels. - Make labels available in all plotting routines. - Subgrid build inside plot_1d (or new function).
Right now nested sampling is tagged onto Swyft externally. It should become part of the core infrastructure.
For details see: https://bestpractices.coreinfrastructure.org
Right now, we determine the left- and right-hand sides of the estimated ratios through coupling the estimators to inputs derived from A or B. Advantages: - Full flexibility in passing...
A method with an interface similar to the plotting routines, but returning mean, mode and credible regions.
In order to calculate Bayes factors, we need information about the prior truncation volume. It would be enough to add a `volume()` method to `RectBoundSampler`, and illustrate the usage in...