pints
pints copied to clipboard
pints-team
Probabilistic Inference on Noisy Time Series
Why? Because it'll be very useful for functionally testing gradient-based methods. I think this can just be copied straight over from that for `GaussianLogPDF`. Whilst we're at it, it might...
During functional testing, I realised there's a mistake in Emcee hammer that means it doesn't converge to the correct target distribution. It should be $q = z^(N-1) p(Y)/p(X)$ where $N$...
Created MultiNest sampler and simplified ellipsoidal nested sampling by creating an Ellipsoid class shared by both samplers. Things done: - Created Ellipsoid class used by both methods - Added MultiNest...
At the moment we've got a class [TunableMethod](https://github.com/pints-team/pints/blob/master/pints/_core.py#L326-L365) that samplers and optimisers and anything else can implement, that provides two methods: ``` n_hyper_parameters -> int set_hyper_parameters(x) ``` where `x` is...
ArviZ provides a simple API to compute the LOO or WAIC for performance assessment of models, see https://arviz-devs.github.io/arviz/api/generated/arviz.waic.html. What this would require however is the pointwise log-likelihood scores of the...
The current implementation of the `pints.NoUTurnMCMC` is at least 10 times slower than `pints.HamiltonianMCMC` on a problem I tried (5 dimensional parameter space, default settings), which probably renders this implementation...
At the moment, setting hyperparameters through the controller uses the internal ._sampler method, which is not ideal.
It seems quite established to compare the distribution of energy transition to the overall energy distribution in a chain to determine whether the momenta / mass matrix where chosen appropriately...
At the moment it's very easy to sample from priors, but we do not have any way to seed these random samples. I find seeding especially useful when exploring the...
As I have banged on about, our job in Pints is partly educational. I created this d3 visualisation of our wish list of the samplers and optimisers which, to me...