scripts

Simulation and Real Data Analysis Scripts

Simulation and real data analysis scripts for the eigenmodel for dynamic multilayer networks. The code was original written to be run on a HPC cluster.

How to Run

The following commands run the simulations or real data analyses and produce the figures in the corresponding section.

Section 4.1 (Parameter Recovery for Varying Network Sizes)

These commands run the simulations and produce Figure 2.

>>> python simulation_parameter_recovery.py
>>> cd output_parameter_recovery/
>>> python process.py
>>> python plot_results.py

Section 4.2.1 (Comparison of Joint and Separate Estimation)

These commands run the simulations and produce Figure 3.

>>> python simulation_joint_seperate_estimation.py
>>> cd output_joint_seperate_estimation/
>>> python plot_results.py

Section 4.2.2 (Comparison of the SMF and MF Variational Inference Algorithms)

The following commands runs the simulations and produce Figure 4.

>>> python simulation_smf_mf_comparison.py
>>> cd output_smf_mf_comparison/
>>> python plot_results.py

To evaluate the run times of the SMF and MF algorithm and produce Figure 5, run the following commands.

>>> python simulation_run_time.py
>>> cd output_run_time/
>>> python plot_results.py

Section 5.1 and Appendix J (ICEWs Networks)

>>> python icews.py

To produce the figures, you will need to run the cells in the corresponding Jupyter notebook:

>>> jupyter notebook output_icews/ICEWS.ipynb

Section 5.2 and Appendix J (Primary School Contact Networks)

>>> python primaryschool.py

To produce the figures, you will need to run the cells in the corresponding Jupyter notebook:

>>> jupyter notebook output_primaryschool/PrimarySchool.ipynb

Appendix H (Evaluation of Information Criteria for Dimension Selection)

The following commands run the simulations and produce Figure 13.

>>> python simulation_dimension_selection.py
>>> cd output_dimension_selection/
>>> python plot_results.py

Appendix I (Additional Simulation: Decreasing sigma as T increases)

The following commands run the simulations and produce Figure 14.

>>> python simulation_parameter_recovery_additional.py
>>> cd output_parameter_recovery_additional/
>>> python process.py
>>> python plot_results.py