pints
pints copied to clipboard
pints-team
Probabilistic Inference on Noisy Time Series
At the moment, it is working on unconstrained prior space.
Again to avoid dependency issues it should be possible (and informative) to recreate the Multinest sampling algorithm in Pints. See [here](https://arxiv.org/abs/0809.3437).
1. Not many changes in PINTS luckily, just replacing a `np.array(x, copy=False)` with `np.asarray(x)` 2. Remaining errors are in `pycma`. These are already fixed in development but we'll need to...
Major numpy version is now available through pip, will likely require some changes
Hi @ben18785 ! We have a hierarchy of methods in the PINTS docs, here: https://github.com/pints-team/pints/blob/master/docs/source/index.rst I think this was a forerunner of your bigger & better interactive tool and the...
Currently the MCMC methods operate in one of two ways: Form 1: ``` def ask(): if self._proposed is None: self._proposed = ... return self._proposed def tell(fx): if self._proposed is None:...
It's the default stopping criterium, so would be nice to see its value during optimisations
Hey all, Recent change to nested sampling changed a method `set_posterior_samples` to `set_n_posterior_samples`. This is more clear, but wondering if it's consistent. For example, (1) there's also a `set_iterations` in...
Most toy models / distributions currently have a single method that tests several things. These should be broken up into proper unit tests - [ ] test_toy_annulus_logpdf.py - [ ]...
Makes it easier to find things