SIS_vs_epsilon2_epsilon3.py

import json
import os

import numpy as np
import xgi

from src.HypergraphContagion import get_opinion_disparity_in_parallel

is_verbose = True
num_processes = len(os.sched_getaffinity(0))

# Epidemic parameters
gamma = 1
tmax = 100
fraction_to_average = 0.1

num_nodes = 10000

num_sims = 1

beta2tilde = 0.2
beta3tilde = 4

output_filename = "empirical_epsilon2_epsilon3_opinion_disparity.json"

with open("Data/SBM/hypergraphs/epsilon_values.json") as file:
    data = json.loads(file.read())
epsilon2 = data["epsilon2"]
epsilon3 = data["epsilon3"]
n = len(epsilon2)
m = len(epsilon3)

data = dict()
arglist = list()

for e2 in epsilon2:
    for e3 in epsilon3:
        fname = f"Data/SBM/hypergraphs/{e2}-{e3}.json"
        H = xgi.read_json(fname, nodetype=int)
        community1 = set(list(H.nodes)[: int(H.num_nodes / 2)])
        community2 = set(list(H.nodes)[int(H.num_nodes / 2) :])
        mean_link_degree = H.nodes.degree(order=1).mean()
        mean_triangle_degree = H.nodes.degree(order=2).mean()

        beta2c = gamma / mean_link_degree
        beta3c = gamma / mean_triangle_degree

        beta2 = beta2tilde * beta2c
        beta3 = beta3tilde * beta3c

        beta = {2: beta2, 3: beta3}
        arglist.append(
            (
                fname,
                gamma,
                beta,
                community1,
                community2,
                tmax,
                fraction_to_average,
                num_sims,
                is_verbose,
            )
        )
print("Simulations started!", flush=True)
psi = get_opinion_disparity_in_parallel(arglist, num_processes)
psi = np.reshape(psi, [n, m], order="C")

data["gamma"] = gamma
data["beta2"] = beta2tilde
data["beta3"] = beta3tilde
data["epsilon2"] = epsilon2
data["epsilon3"] = epsilon3
data["psi"] = psi.tolist()

datastring = json.dumps(data)

with open(f"Data/opiniondisparity/{output_filename}", "w") as output_file:
    output_file.write(datastring)