def run(self, args):
self.error_msg = None
prog = args['prg']
onet = args['onet']
sr = arg_with_default(args, 'sr', DEFAULT_SAMPLE_RATE)
directed = not args['undir']
nodes = arg_with_default(args, 'nodes', DEFAULT_NODES)
edges = arg_with_default(args, 'edges', DEFAULT_EDGES)
gentype = arg_with_default(args, 'gentype', DEFAULT_GEN_TYPE)
print('nodes: {}'.format(nodes))
print('edges: {}'.format(edges))
# load and run generator
gen = load_generator(prog, directed, gentype)
net = gen.run(nodes, edges, sr)
# write net
net.save(onet)
print('done.')
return True
def run(self, args):
self.error_msg = None
prog = args['prg']
out_prog = args['oprg']
sr = arg_with_default(args, 'sr', DEFAULT_SAMPLE_RATE)
directed = not args['undir']
nodes = arg_with_default(args, 'nodes', DEFAULT_NODES)
edges = arg_with_default(args, 'edges', DEFAULT_EDGES)
gentype = arg_with_default(args, 'gentype', DEFAULT_GEN_TYPE)
print('nodes: %s' % nodes)
print('edges: %s' % edges)
# load and run generator
gen = load_generator(prog, directed, gentype)
gen.run(nodes, edges, sr)
# prune and save
gen.prog.dyn_pruning()
gen.prog.write(out_prog)
print('done.')
return True
def run(self, args):
self.error_msg = None
netfile = args['inet']
prog = args['prg']
sr = arg_with_default(args, 'sr', DEFAULT_SAMPLE_RATE)
bins = arg_with_default(args, 'bins', DEFAULT_BINS)
max_dist = arg_with_default(args, 'maxdist', DEFAULT_MAX_DIST)
directed = not args['undir']
runs = arg_with_default(args, 'runs', DEFAULT_RUNS)
gen_type = arg_with_default(args, 'gentype', DEFAULT_GEN_TYPE)
# load net
net = load_net(netfile, directed)
# create base generator
base_generator = create_generator(directed, gen_type)
if base_generator is None:
self.error_msg = 'unknown generator type: %s' % gen_type
return False
# create fitness calculator
# TODO: norm samples configurable
fitness = DistancesToNet(net, get_stat_dist_types(args), bins, max_dist,
norm=Norm.ER_MEAN_RATIO, norm_samples=DEFAULT_NORM_SAMPLES)
fit_maxes = []
fit_means = []
for i in range(runs):
print('run #%s' % i)
gen = load_generator(prog, directed, gen_type)
synth_net = gen.run(len(net.vs), len(net.es), sr)
fit_max, fit_mean, distances = fitness.compute(synth_net)
fit_maxes.append(fit_max)
fit_means.append(fit_mean)
print('fitness (max): %s; fitness (mean): %s' % (fit_max, fit_mean))
print(distances)
mean_fit_max = sum(fit_maxes) / runs
max_fit_max = max(fit_maxes)
min_fit_max = min(fit_maxes)
mean_fit_mean = sum(fit_means) / runs
max_fit_mean = max(fit_means)
min_fit_mean = min(fit_means)
print('\n\n')
print('mean fitness (max): %s; min fitness (max): %s; max fitness (max): %s'
% (mean_fit_max, min_fit_max, max_fit_max))
print('mean fitness (mean): %s; min fitness (mean): %s; max fitness (mean): %s'
% (mean_fit_mean, min_fit_mean, max_fit_mean))
print('done.')
return True
def run(self, args):
self.error_msg = None
prog1 = args['prg']
prog2 = args['prg2']
sr = arg_with_default(args, 'sr', DEFAULT_SAMPLE_RATE)
directed = not args['undir']
nodes = arg_with_default(args, 'nodes', DEFAULT_NODES)
edges = arg_with_default(args, 'edges', DEFAULT_EDGES)
gentype = arg_with_default(args, 'gentype', DEFAULT_GEN_TYPE)
gen1 = load_generator(prog1, directed, gentype)
gen2 = load_generator(prog2, directed, gentype)
gen1.run(nodes, edges, sr, shadow=gen2)
dist1 = gen1.eval_distance
gen2.run(nodes, edges, sr, shadow=gen1)
dist2 = gen2.eval_distance
dist = (dist1 + dist2) / 2
print('eval distance: %s' % dist)
return True
def run(self, args):
self.error_msg = None
prog = args['prg']
sr = arg_with_default(args, 'sr', DEFAULT_SAMPLE_RATE)
directed = not args['undir']
nodes = arg_with_default(args, 'nodes', DEFAULT_NODES)
edges = arg_with_default(args, 'edges', DEFAULT_EDGES)
gentype = arg_with_default(args, 'gentype', DEFAULT_GEN_TYPE)
generator = load_generator(prog, directed, gentype)
generator.run(nodes, edges, sr)
print('is constant? %s' % generator.is_constant())
return True
def run(self, args):
self.error_msg = None
netfile = args['inet']
prog = args['prg']
sr = arg_with_default(args, 'sr', DEFAULT_SAMPLE_RATE)
bins = arg_with_default(args, 'bins', DEFAULT_BINS)
max_dist = arg_with_default(args, 'maxdist', DEFAULT_MAX_DIST)
directed = not args['undir']
runs = arg_with_default(args, 'runs', DEFAULT_RUNS)
gen_type = arg_with_default(args, 'gentype', DEFAULT_GEN_TYPE)
# load net
net = load_net(netfile, directed)
# create base generator
base_generator = create_generator(directed, gen_type)
if base_generator is None:
self.error_msg = 'unknown generator type: {}'.format(gen_type)
return False
# create fitness calculator
# TODO: norm samples configurable
fitness = DistancesToNet(net,
get_stat_dist_types(args),
bins,
max_dist,
norm=Norm.ER_MEAN_RATIO,
norm_samples=DEFAULT_NORM_SAMPLES)
fit_maxes = []
fit_means = []
for i in range(runs):
print('run #{}'.format(i))
gen = load_generator(prog, directed, gen_type)
synth_net = gen.run(net.vcount(), net.ecount(), sr)
distances = fitness.compute(synth_net)
fit_max = max(distances)
fit_mean = mean(distances)
fit_maxes.append(fit_max)
fit_means.append(fit_mean)
print('fitness (max): {}; fitness (mean): {}'.format(
fit_max, fit_mean))
print([stat_type.name for stat_type in fitness.stat_types])
print(distances)
mean_fit_max = sum(fit_maxes) / runs
max_fit_max = max(fit_maxes)
min_fit_max = min(fit_maxes)
mean_fit_mean = sum(fit_means) / runs
max_fit_mean = max(fit_means)
min_fit_mean = min(fit_means)
print('\n\n')
print('mean fitness (max): {}; '
'min fitness (max): {}; '
'max fitness (max): {}'.format(mean_fit_max, min_fit_max,
max_fit_max))
print('mean fitness (mean): {}; '
'min fitness (mean): {}; '
'max fitness (mean): {}'.format(mean_fit_mean, min_fit_mean,
max_fit_mean))
print('done.')
return True