def generate_network_topology(self,
nodes,
link_probability,
weights=None,
types=0):
self.node_size = nodes
if types == 0:
if weights is None:
self.graph_topology = rd.make_random_graph(
nodes, link_probability)
else:
self.graph_topology = rd.make_weighted_random_graph(
nodes, link_probability, weights)
return self.graph_topology
elif types == 1:
return 0
def generate_network_topology(self,
nodes,
link_probability,
wmax,
weights=None,
types=0):
self.node_size = nodes
node_attr = []
if types == 0:
if weights is None:
tp, min, max = rd.make_random_graph(nodes, link_probability)
else:
tp, node_attr, min, max = rd.make_weighted_random_graph(
nodes, link_probability, weights, wmax)
return tp, node_attr, min, max
else:
return 0, 0, 0, 0
ENTROPY_WEIGHT = 0.5 #shold be a placeholder for decaying
EPS = 1e-6
INPUT_FEATURES = 5
EXTRACTED_FEATURES = INPUT_FEATURES * 8
SNODE_SIZE = 100
VNODE_FEATURES_SIZE = 3
NUM_AGENT = 4
ORDERS = 3
GAMMA = 0.99
ALIVE_TIME = 50000
config = tf.ConfigProto(device_count={"CPU": 8},
inter_op_parallelism_threads=1,
intra_op_parallelism_threads=1)
G, node_attr, min, max = rd.make_weighted_random_graph(SNODE_SIZE, 0.5,
"normal", 1)
LAPLACIAN = rd.make_laplacian_matrix(G)
LAPLACIAN_LIST = rd.make_laplacian_list(G, SNODE_SIZE, ORDERS)
LAPLACIAN_TENSOR = np.stack(LAPLACIAN_LIST)
log_dir = "/home/yanzx1993/VNR_A3C_SUMMARY"
#histogram:distribution
#scalar:scalar
def weight_init(shape):
weight = tf.truncated_normal(shape, stddev=0.1)
return tf.Variable(weight)
def bias_init(shape):
import random_graph_gen as rd
import network_topology_gen as nt
# import memory_pool as mp
# import gcn_layer_with_pooling_v2 as gcn
import environment_gen as en
import time
ENTROPY_WEIGHT = 0.5
EPS = 1e-6
INPUT_FEATURES = 5
EXTRACTED_FEATURES = INPUT_FEATURES * 32
SNODE_SIZE = 100
VNODE_FEATURES_SIZE = 3
NUM_AGENT = 4
ORDERS = 3
G = rd.make_weighted_random_graph(SNODE_SIZE, 0.5, "normal")
LAPLACIAN = rd.make_laplacian_matrix(G)
LAPLACIAN_LIST = rd.make_laplacian_list(G, SNODE_SIZE, ORDERS)
LAPLACIAN_TENSOR = np.stack(LAPLACIAN_LIST)
def weight_init(shape):
weight = tf.truncated_normal(shape, stddev=0.1)
return tf.Variable(weight)
def bias_init(shape):
bias = tf.constant(0.1, shape=shape)
return tf.Variable(bias)