mackelab / lfi-experiments Goto Github PK

seed_obs should become part of the abstract base class likelihoodfree/Simulator.py, rather than being part of its descendants (not urgent for now)

Todo:

• move https://github.com/mackelab/likelihoodfree-glm to this repo
• adapt API to match likelihoodfree.Inference
• delete original repo once everything works

Won't work on this for now, but code for the Lotka-Volterra example based on the epsilonfree implementation is in a feature branch. Before merging into master we'd need to: Give credit for the original implementation, and adapt lv.py to match the inference API of likelihoodfree.

In the future (not urgent for now) we'll need to fix up run.py to be more user friendly:

• especially sim_kwargs and train_kwargs need rethinking
• passing strings required ", should also work without or at least with ', something like observe=\"weights\" is awkward

Your repo does not comply to the standards we defined in the lab.
Make sure for your repo to have:

• a description including the github handle of the owner
• a > 3 line README.md

If you don't update your repo, it will be disabled and then archived.

Todo @jan-matthis:

• move https://github.com/mackelab/likelihoodfree-neurons (work in progress)
• adapt API to match likelihoodfree.Inference (work in progress)
• delete https://github.com/mackelab/likelihoodfree-neurons
• make sure Cython version can be dropped in for faster simulation
• hh.py got cython flag which will change import, just need to make sure that interface is the same and seeding works properly
• add install instructions to setup.py and README.md

Todo @ppjgoncalves and @jan-matthis:

• make the implementation is consistent with https://github.com/mackelab/likelihoodfree-HH (i.e., the python 2 version)
• move genetic algorithm code, will become part of lfmods/hh.py
• resolve allendb issue with python 2, see allen_db_issue.pdf
• delete https://github.com/mackelab/likelihoodfree-HH once all is done