simulate and fit reinforcement learning models in behavioural experiment tasks
best pronounced "simfitterly" or, if you find it funnier, "simfit...really!?"
simfitRLy is the main library used in my phd thesis, On State Representations and Behavioural Modelling Methods in Reinforcement Learning
this code has not yet been fully tested and packaged up nicely. use at your own risk. until i've had time to fix it up properly and understand what license will be the best fit, this repo is (c) Henrik Siljebråt.
we recommend using anaconda, specifically mambaforge, a much faster version of the default conda tool which also uses conda-forge channel as default.
create a new conda environment and install required libraries using the included environment file:
mamba env create --file environment.yml