def calculate(): # Выполнить необходимые расчеты и построить графики
# Задать конфигурацию на основе ввода пользователя
m = [int(n) for n in configuration.field_input_configuration.get().split()]
graph.Vertex.max_requests = m
graph.Vertex.number_of_types = len(graph.Vertex.max_requests)
# Основные параметры для расчета производительности
TO = program.TO = [float(n) for n in configuration.field_TO.get().split()
] # Время выполнения операций
τ = program.τ = [float(n) for n in configuration.field_τ.get().split()
] # Время цикла оперативной памяти
ξ = program.ξ = float(configuration.field_ξ.get()) # Коэффициент связности
# Дополнительные параметры для расчета производительности (вычисляются из основных)
ν = program.ν = [ξ / (TO + ξ * τ)
for TO, τ in zip(TO, τ)] # Параметр интенсивности
μ = program.μ = [1 / τ for τ in τ] # Параметр обслуживания
program.ρ = [ν / μ for ν, μ in zip(ν, μ)]
# Построить граф
graph.plot_graph(m)
# Построить график зависимости производительности от количества процессоров
plt_data = program.plot_graph_for(
m,
int(plot_data.field_input_plot_for.get()) - 1,
int(plot_data.field_plot_up_to.get()))
# Вывести статистику в окно
txt = ''
for m, q, p in zip(plt_data[0], plt_data[1], plt_data[2]):
txt += ('{:<10} {:<34} {:<30}\n'.format(m, q, p))
output_window.text.config(state='normal')
output_window.text.delete(1.0, 'end')
output_window.text.insert('end', txt)
output_window.text.config(state='disabled')
def main():
print("\nObjective Genetic Algorithm\n")
if len(sys.argv) != 11 and len(sys.argv) != 2:
print("Wrong number of parameters (expected 2 or 11) but got " +
str(len(sys.argv)))
exit()
if len(sys.argv) == 2:
print("Number of tasks: ", end="")
nTasks = int(input())
print("Number of machines: ", end="")
nMachines = int(input())
print("Size of MAKESPAN population: ", end="")
populationSize = int(input())
print("Mutation factor (percentage): ", end="")
mutationFactor = int(input())
print("Crossover factor (percentage): ", end="")
crossoverFactor = int(input())
print("Elitism factor (percentage): ", end="")
elitismFactor = int(input())
print("Crossover Operator (1..4): ", end="")
crossoverOperator = int(input())
print("Mutation Operator (1..2): ", end="")
mutationOperator = int(input())
print("\nNumber of executions: ", end="")
executions = int(input())
else:
nTasks = int(sys.argv[2])
nMachines = int(sys.argv[3])
populationSize = int(sys.argv[4])
mutationFactor = int(sys.argv[5])
crossoverFactor = int(sys.argv[6])
elitismFactor = int(sys.argv[7])
crossoverOperator = int(sys.argv[8])
mutationOperator = int(sys.argv[9])
nExecutions = int(sys.argv[10])
filename = sys.argv[1]
iterations = 50000
summaryLogLines = []
makespans = []
flowtimes = []
averages = []
for i in range(nExecutions):
arguments = ' '.join(sys.argv[1:])
print("EXECUTION " + str(i + 1) + " of " + arguments)
if crossoverOperator == 1:
if mutationOperator == 1:
bestIndividual, executionGenerations, executionAverages, executionMakespans, executionFlowtimes = genetic_algorithm11.GA(
filename, nTasks, nMachines, mutationFactor,
crossoverFactor, elitismFactor, populationSize,
crossoverOperator, mutationOperator, iterations).execute()
else:
bestIndividual, executionGenerations, executionAverages, executionMakespans, executionFlowtimes = genetic_algorithm12.GA(
filename, nTasks, nMachines, mutationFactor,
crossoverFactor, elitismFactor, populationSize,
crossoverOperator, mutationOperator, iterations).execute()
elif crossoverOperator == 2:
if mutationOperator == 1:
bestIndividual, executionGenerations, executionAverages, executionMakespans, executionFlowtimes = genetic_algorithm21.GA(
filename, nTasks, nMachines, mutationFactor,
crossoverFactor, elitismFactor, populationSize,
crossoverOperator, mutationOperator, iterations).execute()
else:
bestIndividual, executionGenerations, executionAverages, executionMakespans, executionFlowtimes = genetic_algorithm22.GA(
filename, nTasks, nMachines, mutationFactor,
crossoverFactor, elitismFactor, populationSize,
crossoverOperator, mutationOperator, iterations).execute()
else:
bestIndividual, executionGenerations, executionAverages, executionMakespans, executionFlowtimes = genetic_algorithm41.GA(
filename, nTasks, nMachines, mutationFactor, crossoverFactor,
elitismFactor, populationSize, crossoverOperator,
mutationOperator, iterations).execute()
makespans.append(executionMakespans)
flowtimes.append(executionFlowtimes)
averages.append(executionAverages)
graphName = '-'.join(sys.argv[1:]).replace(
".txt", "") + "-" + str(i + 1) + "-graph.png"
print(graphName)
graph.plot_graph(graphName, executionGenerations, executionMakespans,
executionAverages, graphName)
summaryLogLine = logLine(filename, bestIndividual, executionAverages,
populationSize, crossoverOperator,
crossoverFactor, mutationOperator,
mutationFactor, elitismFactor)
print(summaryLogLine)
summaryLogLines.append(summaryLogLine)
logFilename = filename.replace(".txt", "_log.txt")
logAverages = []
logMakespans = []
for i in range(len(executionGenerations)):
averageSum = 0.0
makespansSum = 0.0
for j in range(nExecutions):
averageSum += (averages[j])[i]
makespansSum += (makespans[j])[i]
logAverage = averageSum / nExecutions
logMakespan = makespansSum / nExecutions
logAverages.append(logAverage)
logMakespans.append(logMakespan)
graphName = '-'.join(sys.argv[1:]).replace(".txt", "") + "-graph.png"
graph.plot_graph(filename, executionGenerations, logMakespans, logAverages,
graphName)
logSummary(summaryLogLines)
for n2 in nodes2:
if np.random.rand() < p_inter:
G.add_edge(n1, n2)
has_inter_edge = True
if not has_inter_edge:
G.add_edge(nodes1[0], nodes2[0])
#print('connected comp: ', len(list(nx.connected_component_subgraphs(G))))
G = G.to_directed()
G = gl.networkx2graph(G)
return G
def max_n_nodes(graphs):
max_nodes = 0
for graph in graphs:
n_nodes = len(graph.nodes)
if n_nodes > max_nodes:
max_nodes = n_nodes
return max_nodes
if __name__ == "__main__":
gnf_dataset_keys = [
'ego_small', 'community_small', 'graph_rnn_protein', 'ego', 'grid',
'community', 'community_medium'
]
dataset = DatasetErdosRenyiNodes(partition='test', overfit=True)
for graph in dataset.graphs:
gl.plot_graph(graph)
from graph import author_graph, plot_graph
from submissions import count_deltas, submission_authors
REDDIT = Reddit(client_id=client_id,
client_secret=client_secret,
user_agent=user_agent)
cmv = REDDIT.subreddit('changemyview')
for submission in cmv.hot(limit=5):
print('\n')
print(submission.url)
nDelta = count_deltas(submission, REDDIT)
print(nDelta, 'deltas')
authorSubs = submission_authors(submission)
G = author_graph(authorSubs)
plot_graph(G)
partition = best_partition(G)
nGroups = len(set(partition.values()))
print(nGroups, 'group(s)')
distance_matrix, static_connection_graph)
punishment_matrix, transport_matrix = graph.merge_graph_to_matrix(
punishment_graph)
print(
f'{np.count_nonzero(~np.isnan(punishment_matrix))} / {nbr_cities*nbr_cities} connections'
)
#travel_time_graph, punishment_graph, score_graph = graph.generate_punishment_graph_from_distance(nbr_transport_types, distance_matrix, city_extra_points)
# Set arguments
args = (punishment_matrix, transport_matrix, city_extra_points)
kwargs = {
'start_city': 2,
'target_city': 9,
'nbr_ants': 50,
'verbose': True,
'evaporation': 0.5,
'alpha': 1.0, # pheromones
'beta': 3.0 # scores
}
# Run colonies in parallel
best_path, best_score, all_results = aco.run_parallel_colonies(
nbr_parallel_jobs, nbr_colonies, args, kwargs)
# Run 1 colony
#best_path, best_score = aco.summon_the_ergodic_colony(*args, **kwargs)
# Plot the graph
graph.plot_graph(city_locations, punishment_graph, best_path)
def graph(self):
plot_graph()
'community_overfit'
'erdosrenyinodes_0.25_overfit'
:return:
'''
if dataset_name.startswith("erdosrenyinodes"):
_, p, overfit = dataset_name.split("_")
dataset = d_creator.DatasetErdosRenyiNodes(p=float(p), partition=partition, overfit=overfit=="overfit")
elif dataset_name.startswith("erdosrenyi"):
_, n_samples, n_nodes, n_edges = dataset_name.split("_")
dataset = d_creator.DatasetErdosRenyi(None, int(n_nodes), int(n_edges), partition, directed)
elif dataset_name == "community_ours":
dataset = d_creator.DatasetCommunity(partition=partition)
elif dataset_name == "community_overfit":
dataset = d_creator.DatasetCommunity(n_samples=100)
else:
raise Exception("Wrong dataset %s" % dataset_name)
return dataset
if __name__ == "__main__":
datasets = ['erdosrenyinodes_0.25_none']
for dataset in datasets:
print(dataset)
dataset = retrieve_dataset(dataset, "test")
for sample in dataset.graphs:
graph.plot_graph(sample)
ps.author_id(data, 'aris anagnostopoulos')
# Given the author's ID, find her/his publications:
ps.publications_for_author_id(data, 256176)
# Create the set of all author IDs in order to build the graph:
list_of_authors_IDs = ps.return_all_the_authors_IDs(data)
len(list_of_authors_IDs)
# Graph:
G = g.create_graph(list_of_authors_IDs)
g.plot_graph(G)
# We need a dictionary containing all the publications for a given author
# in order to give a relationship author-publication:
authorsDict = {}
for i in range(len(data)):
for author in data[i]['authors']:
for j in range(len(data[i]['authors'])):
key = data[i]['authors'][j]['author_id'] # the key is the ID for every author
if key in authorsDict.keys():
if data[i]['id_publication_int'] not in authorsDict[key]:
authorsDict[key].append(data[i]['id_publication_int'])
else:
authorsDict[key] = [data[i]['id_publication_int']]
authorsDict
def entry_point():
name = request.args.get('lname')
yvals, xvals = get_stock_value(name)
plot_graph(app, xvals, yvals, name)
return redirect('/graph')