pints
pints copied to clipboard
pints-team
Probabilistic Inference on Noisy Time Series
I think it'd be good to have suggested bounds on hyperparameters to help people to choose sensible values. This may also help the optimisation-based approach to selecting hyperparameters (by putting...
https://www.python.org/dev/peps/pep-0257/#what-is-a-docstring > Multi-line docstrings consist of a summary line just like a one-line docstring, followed by a blank line, followed by a more elaborate description. The summary line may be...
At the moment we've got both `set_iterations` and `set_n_parameters`. Would be good to have a convention and stick to it :-) @martinjrobins @ben18785 @chonlei ?
https://www.cs.ox.ac.uk/projects/PINTS/functional-testing/#nestedbanananestedellipsoidsampler
- [ ] MALA - [ ] Relativistic MC - [ ] Hamiltonian MCMC @MichaelClerx thanks for pointing these out!
Suppose one chain is stuck on one mode; another on another mode. If those two chains may sample independently from each mode, the ESSs will be high when, really, they...
E.g. - Seed - Run 100 iterations - Check that there's sufficient change within those iterations (and reduce n if possible) - Store output, either in CSV or in code...
Is allowed to be just an output value frozen in place, provided we check it at the time.
See the surface plot example. Not sure why this is, but probably isn't desirable