You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
@agarbuno: This PR is based on your EKS implementation. However, the EKS update needs some revision: First, it errored due to dimensional mismatches, which led me to look at the implicit equation implemented in lines 250-256 in src/EKP.jl. That equation seems to differ from equation (2.8a) in the Calibrate-Emulate-Sample paper, and so I didn't know how to correctly adjust it. Would be great if you could have a look at it!
The current implementation runs through (thanks to replacing the + in the beginning of line 253 with a .+), but it doesn't produce a correct EKS iteration (I ran the EKS test case implemented in test/EKP/runtests.jl and plotted the evolution of the particles in each iteration -- the particles are diverging away from the initial ensemble rather than being distributed in regions of high posterior probability).
76: Change code infrastructure to accommodate both EKI and EKS r=bielim a=bielim
This PR is an interim solution, to be superseded by PR #70 as soon as the EKS implementation has been finalized. It implements the new "EK infrastructure" with the abstract type `Process` that allows the user to choose between an EK `Sampler` and an EK `Inversion`, but since the correct ensemble update for the EK Sampler is not known yet (see discussion in the comment section of PR #70), an attempt to instantiate an EK `Sampler` is prohibited and will throw an error.
The point of merging this PR before completion of the EK Sampler implementation is to not block the further development of the code base, which depends on the infrastructure introduced in this PR.
Co-authored-by: Melanie Bieli <melanie.bieli@bluewin.ch>
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This PR adds a Julia implementation of an Ensemble Kalman Sampler. It supersedes PR #41.