Code for Cone, Clopath and Shouval 2024 "Learning to Express Reward Prediction Error-like Dopaminergic Activity Requires Plastic Representations of Time"
The code for the main FLEX model is located in \main_model.
"Main_vectorized_simple_dopa.m" includes the main code for the simulations - running the program will run a demo simulation and plot relevant figures afterwards. Comments are included in the file for further instruction.
"spiking_parameters_simple_dopa.m" includes the parameters for use in the main file.
The code for data analysis is located in \data_analysis.
Data from Amo et al. 2022 is included in \data_analysis. Data from Coddington and Dudman 2018 is available from the authors upon request.
The code for the versions of the model with multiple cues and value encoding is located in \multiple_cues_and_value_model.
It is structured like the main_model folder, so running "Main_vectorized_simple_dopa_m_c.m" will run a demo simulation with distractor cues and plot relevant figures afterwards. Comments are included in the file for further instruction.
"spiking_parameters_simple_dopa.m" includes the parameters for this simulation, as before.
To test the value network, run "value_wrapper.m"
This code was written and run in MATLAB 2023b. Please direct any questions to ianconehed(at)gmail(dot)com. It was tested both on Mac and Windows platforms. It should take no more than 30 minutes to run the code.