pip install numpy scipy matplotlib seaborn pandas jupyter nengo nni
- model.ipynb contains code for simulating individual trials and plotting dynamics
- analysis.ipynb contains code for simulating many trials, for running multiple experiments with different parameters, and for plotting comparisons to the empirical data
- nni/compare_mechanisms.py contains the core optimization code for fitting model parameters to the human data
- nni/experiment.py is the wrapper code for starting multiple nni experiments simultaneously: edit the file to specify the hyperparameters, then run
python experiment.py - nni/collect.py is used to collect the data generated by running experiment.py, once all nni experiments are complete. Just run
python collect.py, which will spit out a .json file that contains the best-fit parameters for each individual. Then use analysis.ipynb to simulate the plot the results.