ultranest import
Closed this issue · 0 comments
Is your feature request related to a problem? Please describe.
In case you want to support imports from ultranest, here are some pointers:
In the user-provided folder (for storing or resuming), ultranest leaves `results/points.hdf5``, a file storing all sampled points.
It uses a highly space-efficient packing based on hdf5, which can be opened easily with h5py.
The table points
contains the sampled points; the columns are:
- loglikelihood threshold under which the point was sampled,
- loglikelihood of the point,
- a quality indicator (0 for MLFriends, otherwise the number of steps in the step sampler),
- u-space (unit cube) coordinates,
- p-space (transformed parameters) coordinates.
This is documented here: https://johannesbuchner.github.io/UltraNest/performance.html#output-files
Looking at the ultranest.integrator.read_file
function may be helpful (docs: https://johannesbuchner.github.io/UltraNest/ultranest.html#ultranest.integrator.read_file, source: https://johannesbuchner.github.io/UltraNest/_modules/ultranest/integrator.html#read_file).
Describe the solution you'd like
Read in ultranest results with anesthetic.
Describe alternatives you've considered
ultranest already prints out where the live points are distributed during the run, performs the U-test diagnostic, and, finally, provides posterior distributions as corner plots.
Additional context
Hi 👋