107 lines
2.8 KiB
Python
107 lines
2.8 KiB
Python
import multiprocessing
|
|
import os
|
|
import requests
|
|
|
|
|
|
def worker(outqueue, id):
|
|
print(f"Worker {id} is preparing...")
|
|
import numpy as np
|
|
import networkx as nx
|
|
|
|
with open("/dkebab/GBA1000.txt") as f:
|
|
n = int(f.readline()) # pontok száma
|
|
m = int(f.readline()) # élek száma
|
|
A = [[0] * n for _ in range(n)] # táblázat, ahol az élek helyén 1 van, ammeg 0
|
|
for _ in range(m):
|
|
[u, v] = map(int, f.readline().split())
|
|
A[u][v] = 1
|
|
A[v][u] = 1
|
|
|
|
input_graph = nx.Graph(np.array(A))
|
|
|
|
def max_comp_size(G):
|
|
return max([len(c) for c in nx.connected_components(G)])
|
|
|
|
predelete_count = int(id / 3)
|
|
target = len(input_graph.nodes) / 2
|
|
|
|
if predelete_count:
|
|
print(f"Worker {id} always deletes the {predelete_count} highest degree nodes")
|
|
|
|
print(f"Worker {id} is working...")
|
|
while True:
|
|
del_list = []
|
|
G = input_graph.copy()
|
|
# A fresh new start
|
|
|
|
for _ in range(predelete_count):
|
|
highest_node = None
|
|
highest_node_deg = 0
|
|
for node in G.nodes:
|
|
if G.degree[node] > highest_node_deg:
|
|
highest_node = node
|
|
highest_node_deg = G.degree[node]
|
|
|
|
del_list.append(highest_node)
|
|
G.remove_node(highest_node)
|
|
|
|
while max_comp_size(G) > target:
|
|
u = np.random.choice(G.nodes)
|
|
if u not in del_list:
|
|
del_list.append(u)
|
|
G.remove_node(u)
|
|
|
|
outqueue.put(del_list)
|
|
|
|
|
|
def worker_watchdog(outqueue, id):
|
|
while True:
|
|
try:
|
|
worker(outqueue, id)
|
|
except KeyboardInterrupt:
|
|
break
|
|
except:
|
|
print(f"Worker {id} died! Restarting...")
|
|
|
|
print(f"Worker {id} exited!")
|
|
|
|
|
|
def denumpy_result(numpy_type) -> list:
|
|
denumyed = []
|
|
for i in numpy_type:
|
|
denumyed.append(int(i))
|
|
|
|
return denumyed
|
|
|
|
|
|
def main():
|
|
target_url = os.environ['TARGET_URL']
|
|
multiprocessing.set_start_method('spawn')
|
|
outqueue = multiprocessing.Queue()
|
|
processes = []
|
|
|
|
for id in range(os.cpu_count()):
|
|
p = multiprocessing.Process(target=worker_watchdog, args=(outqueue, id))
|
|
p.start()
|
|
processes.append(p)
|
|
|
|
local_best = 1000
|
|
|
|
while True:
|
|
result = outqueue.get()
|
|
result_score = len(result)
|
|
|
|
if result_score < local_best:
|
|
print(f"New local best found: {local_best} -> {result_score}")
|
|
local_best = result_score
|
|
|
|
try:
|
|
r = requests.post(target_url, json=denumpy_result(result))
|
|
r.raise_for_status()
|
|
except (ConnectionError, requests.HTTPError) as e:
|
|
print(f"Error while posting result: {e}")
|
|
|
|
|
|
if __name__ == '__main__':
|
|
main()
|