This repository contains code to generate and analyse simulated data that is used in the study of Frequency-sensitivity and magnitude-sensitivity in decision-making: predictions of a theoretical model-based study which is currently under review.
Simulation code provided in this repository has been developed by Thomas Bose as part of the DiODe project (https://diode.group.shef.ac.uk/) and is open source. It may be used and modified for non-commercial use.
Simulations and data analysis were performed in Python code (Python 3) using Jupyter notebooks. For more information on Jupyter notebooks including installation instructions please visit http://jupyter.org/.
A major part of the data generation notebooks provided uses ipyparallel to run calculations in parallel in a standard Python 3 environment. Information on how to install and use the ipyparallel package can be found following this link https://ipyparallel.readthedocs.io/en/latest/.
Jupyter notebooks for data generation were run using runipy in a script on a Linux Ubuntu machine. More information on runipy can be found here. First, the script activate.sh (which is provided in the data generation folders) needs to be run to activate a suitable conda environment (in this case it is called python3) and then four engines are activated for the use with ipyparallel. Running activate.sh will open a new terminal window. Executing run_nb_All.sh (also provided) in the new window then runs all notebooks one after another in the order given in the script file. Notebook output is stored in a separate file as specified in run_nb_All.sh. Of course, both scripts should be modified as required.
Some of the calculations take a significant amount of time (~ several hours or days). Therefore we also provided the data files produced by the code which are placed in subfolders called DataFiles. Notebooks to produce all graphics are also included in the data analysis folders. Images will be plotted inside the notebooks and may be exported using standard Matplotlib functionality. The Figures subfolder which can be found in the DataAnalysis folder contains examples of exported figure files.