mackelab/mnle-for-ddms

Research code for Mixed Neural Likelihood Estimation (MNLE, Boelts et al. 2022)

mnle-for-ddms's Issues

Include value of experimental conditions (e.g. stimulus) in the training and posterior sampling for MNLE.

Greetings,

Firstly, I would like to express my appreciation for this powerful tool.
Currently, I am trying to fit a model (DDM) with three experimental conditions (e.g. stimulus) and 16 parameters. My problem is that I do not know how to sample from the posterior distributions considering these conditions. For the training, I just include them as additional inputs to the network, as suggested in the paper.
Then, should I consider them as observed data (and also as inputs for the training)?

Thanks in advance,
Alex

Integrate MNLE into the sbi toolbox

This is will completed soon, see sbi-dev/sbi#638

[Code Quality] Incomplete Repo information

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.

Unable to get ddm task from sbibm library

In notebooks/Paper-Figure-2-likelihood-comparison.ipynb. There is a line task = sbibm.get_task("ddm") using the latest version of sbibm v1.1.0. With two_mooons as param, it works.

There is no ddm task in the sbibm tasks folder - https://github.com/sbi-benchmark/sbibm/tree/main/sbibm/tasks