Computer codes for "Network games with heterogeneous players"

These computer codes contain valuable information necessary to reproduce the findings presented in the paper "Network games with heterogeneous players". We generated CRS and HP networks with arbitrary homophily indices using MATLAB scripts (MATLAB R2023b), and implemented evolutionary simulations of multi-agent systems in Python 3.13.5 (Anaconda, 64-bit Windows). The usage of all scripts is documented inline. Description of Computer code: “CRS_Network_Generation.m” This MATLAB script implements a method for generating an Erdős-Rényi-like network with an island model structure, specifically designed for CRS (i.e., conformists, rebels, and stubborn agents) networks. The code first calculates connection probabilities based on predefined proportion parameters and a target average network degree, then constructs the network adjacency matrix while ensuring the preservation of node type proportions. Outputs include calculated probability terms, homophily indices of the generated network, and CSV files of nodes and edges for further analysis. Description of Computer code: “HP_Network_Generation.m” This MATLAB script generates a HP network derived from a pre-existing CRS network (generated by the companion script "CRS_Network_Generation.m"). The core functionality involves augmenting each node in the CRS network with stochastic payoffs (columns epsilon_0 and epsilon_1) that follow uniform random distributions, as specified in SI Appendix, section S2B. The script outputs an enhanced node attribute file with original CRS network structure and new stochastic payoff attributes. Description of Computer code: “ODE_Core.m” This MATLAB script implements the ODE equations underlying the evolutionary dynamics of network CRS games, and is specifically designed to generate the basin of attraction plots featured in the main text figures. The code integrates the formulation of the ODEs, numerical integration, and vector field visualization to reproduce the dynamical behaviors of CRS games. Description of Computer code: “CRS_Network_Games.ipynb” This Python script implements the agent-based simulation of evolutionary dynamics for network CRS games, and is specifically designed to characterize the time evolution of strategy proportions. The code integrates network construction from CSV inputs, probabilistic strategy update (individual update probability and information probability), and real-time visualization of dynamical outcomes to reproduce the evolutionary behaviors of CRS games. Description of Computer code: “HP_Network_Games.ipynb” This Python script implements the agent-based simulation of evolutionary dynamics for network HP games. Similar to "CRS_Network_Games.ipynb", the key distinction lies in the use of heterogeneous individual payoff functions.