Raw data of the heterogeneous Hegselmann-Krause model on network ensembles

# Raw data of the heterogeneous Hegselmann-Krause model on network ensembles
This is the raw data underlying the results of the article *«On the effects of over-compromising: heterogeneity and network effects on a bounded confidence opinion dynamics model.»*.

For each measured combination of the parameters, there is one gzipped file. The parameters are:

 - Lower and upper bounds of the confidence interval, [ε_l, ε_u].
 - Topology: the different types of networks and the average degree with which the networks are generated.
 - System size
 - Number of realizations for the parameter combination
 
The single files follow a naming scheme of `data_HK_uni[{eps_l},{eps_u}]_topo={topology}_N={N}_trajrecord=0_{m}real.dat.gz`,  where:

 - `{eps_l},{eps_u}` are the values of the lower and upper bounds of the confidence interval.
 - `{topology}` contains the type of network and the average degree. The possibilities are `BA_k=10`, `ER_c=10`, `sl1`, `sl2`, and `sl3`.
 - `N` is the system size. The sizes are powers of two.
 - `trajrecord=0` signals the fact that file contains only the final state.
 - `{m}` is the number of realizations.

# Data format
Each file contains the final state of each realization back to back. Each final state is encoded as three lines:

 - The convergence time is a single integer with a line prefix '\# iterations:'
 - The positions of all clusters in opinion space with a line prefix '\# ' (unsorted)
 - The number of agents in each of the clusters without a line prefix

# Folders structure
The files are organized as follows:

 - **`phase_plots.tar`**: contains the data for the different phase plots (full exploration of the [ε_l, ε_u] space) with `N=16384` and `m=100` realizations.
     - **`ER`** contains the data for Erdos Renyi with mean degree of 10 (c=10)
     - **`BA`** contains the data for Barabasi Albert with a mean degree of 10 (k=10)
     - **`SL`** contains the data for Square lattice with first, second and third nearest neighbors (k=4, 8, 12)
 - **`swipes.tar`** contains the data for the finite size effects study at fixed ε_l with `m=1000` realizations.
     - **`ER`** contains the data for Erdos Renyi with mean degree of 10 (c=10) with ε_l = 0.05
     - **`BA`** contains the data for Barabasi Albert with a mean degree of 10 (k=10) with ε_l = 0.05
     - **`SL`** contains the data for Square lattice with third nearest neighbors (k=12) with ε_l =0.03
 - the different videos referenced in the main text and the SM follow various naming schemes:
     - **`scatter3D_el_eu_Smax_uni_{topology}_N=16384.mp4`**: 360° rotation of the 3D visualisation of the data leading the average phase plots.
     - **`scatter2D_el={eps_l}_eu_Smax_extremism_{topology}_SizeEffect.mp4`**: evolution of the scatter plot leading the finite size study as a function of N.
     - **`scatter3D_el={eps_l}_eu_Smax_extremism_{topology}_SizeEffect.mp4`**: same as before, but in 3D where the Z-axis is the extremism.
     - **`scatter_x0_xt_{topology}_N={N}_{realization_type}.mp4`**: time evolution of the scatter plot of the opinion at time `t` versus initial opinion, color-coded with the extremism. {realization_type} can be mild, skewed or U-turn.
     - **`traj_2D_SL_k=12_N=16384_{realization_type}.mp4`**: because of the spatial embedding, the time evolution of those realizations on the Square Lattice can be visualized in 2D.

# Python example for reading the format
An example script, which visualizes <S\> vs ε_u graph for the largest size of the ER case, with a function to read this format is given in `example.py`.