/CLSNA-2Party-Polarization

Primary LanguageR

CLSNA-2Party-Polarization

This repository contains code and data supporting the article "Disentangling positive and negative partisanship in affective polarization using a coevolving latent space network with attractors model":

https://arxiv.org/abs/2109.13129

Data

Data used for analysis are 1) Twitter congressional hashtag networks; 2) Reddit comment networks.

They are stored as R binary files (.RData), available in Data/.

Code

The corresponding code to conduct both simulation and analyses on real data sets is provided in Code/.

  • Code/model/: code for implementations of model and inference with no change-point.

  • Code/model_change_point/: code for implementations of model and inference with multiple change-points.

  • The main files for simulation study with synthetic data are Code/model/main_simulation.R and Code/model_change_point/main_simulation_mc.R.

  • Figure 1,2: Code/app_plot_density.R.

  • Figure 4:Code/sim_nets_traces.R.

  • Figure 5, 7, Table 1,2: Code/model/main_application.R, Code/model/posterior_plot_app.R.

  • Figure 6, 8: Code/model_change_point/main_application_mc.R, Code/model_change_point/plot_sub_window_1cp.R.

To report bugs encountered in running the code, please contact Xiaojing Zhu at xiaojzhu@bu.edu