def make_planar(G):
if planarity.is_planar(G.edges()):
return G
all_kuratowski_edges = []
next_G = G.copy()
while planarity.is_planar(G.edges()) == False:
edge_list = planarity.kuratowski_edges(G)
print(len(edge_list))
G.remove_edges_from(edge_list)
for edge in edge_list:
all_kuratowski_edges.append(edge)
itr = 0
replaced = []
for edge in all_kuratowski_edges:
neighbors = nx.neighbors(G, edge[1])
if itr < len(neighbors) and edge not in replaced:
G.add_edge(edge[0], neighbors[itr])
elif itr >= len(neighbors) and edge not in replaced:
global neighbors
neighbors = nx.neighbors(G, edge[0])
itr = 0
itr = itr + 1
print(planarity.is_planar(G), nx.number_of_edges(G))
print(planarity.is_planar(next_G), nx.number_of_edges(next_G))
def round_robin_thickness_graphs(g):
"""
Generate round robin thickness graphs
Take best known solution (naive)
Allocate n-1 graphs (if > 1 in naive solution)
Sort Edges
Round robin edge assignment
(i.e. try to spread out the edges for each node across the graphs)
:param g:
:return:
"""
naive_best = naive_thickness_graphs(g)
if len(naive_best) == 1:
return naive_best
gs = [nx.Graph() for _ in range(len(naive_best) - 1)]
gedges = [e for e in g.edges()]
gedges.sort()
vs = set()
for e in g.edges():
vs.add(e[0])
vs.add(e[1])
for v in vs:
for cur in gs:
cur.add_node(v)
not_added = []
for i, edge in enumerate(gedges):
gi = i % len(gs)
gs[gi].add_edge(edge[0], edge[1])
if not is_planar(gs[gi]):
not_added.append(edge)
gs[gi].remove_edge(edge[0], edge[1])
not_added_again = []
for e in not_added:
added = False
for gcur in gs:
gcur.add_edge(e[0], e[1])
if is_planar(gcur):
added = True
break
else:
gcur.remove_edge(e[0], e[1])
if not added:
not_added_again.append(e)
if len(not_added_again) > 0:
return naive_best
return gs
def thickness(G):
# Complete Graphs
#
# The thickness of an arbitrary complete graph is given by
# floor((n+7)/6) (except for where n=9,10 where thickness is 3)
#
# Alekseev, V. B.; Goncakov, V. S.
# https://mathscinet.ams.org/mathscinet-getitem?mr=0460162
if G.number_of_edges() == edge_count_of_complete_graph(
G.number_of_nodes()):
if G.number_of_nodes() == 9 or G.number_of_nodes() == 10:
return 3
else:
return int(floor((G.number_of_nodes() + 7) / 6))
# Planar Graphs
#
# The thickness of any planar graph is 1
if is_planar(G):
return 1
# Small Graphs
#
# If a graph is pretty small, we can run a brute-force-search
if G.number_of_edges() < 8:
return brute_force_thickness(G)
# All other graphs
#
# since we couldn't find a characterization that helps with this graph, we check using
# Alex's brute-force-like search algorithm
#
# note: this is not guaranteed to yield the actual thickness; only a number greater than or equal to it
return round_robin_thickness(G)
def naive_thickness_graphs(g):
"""
Generate the thickness graphs themselves
:param g:
:return:
"""
vs = set()
gs = [nx.Graph()]
for e in g.edges():
vs.add(e[0])
vs.add(e[1])
added = False
for current in gs:
current.add_edge(e[0], e[1])
if is_planar(current):
added = True
break
else:
current.remove_edge(e[0], e[1])
if not added:
ng = nx.Graph()
ng.add_edge(e[0], e[1])
gs.append(ng)
for g in gs:
for v in vs:
g.add_node(v)
return gs
def pmfg(df):
n = len(df)
A = np.zeros([n,n])
rho = df.as_matrix()
rholist = []
for i in range(0, n):
for j in range(i+1, n):
rholist.append([rho[i][j], i, j])
rholist = np.array(rholist)
rholist = rholist[rholist[:,0].argsort()].tolist()
rholist = [rholist[len(rholist)-i-1] for i in range(len(rholist))]
control = 0
edgelist = []
for t in range(0, len(rholist)):
if control <= 3*(n-2)-1:
i, j = rholist[t][1], rholist[t][2]
A[i][j], A[j][i] = 1, 1
edgelist.append((str(i),str(j)))
if is_planar(edgelist) == False:
A[i][j], A[j][i] = 0, 0
edgelist = edgelist[:-1]
else:
control += 1
return pd.DataFrame(data=A, columns=df.columns, index=df.index)
def construct_pmfg(self,
df_distance,
matrix_panel,
inv=False,
threshold=0.0):
"""Trim the provided adjacency matrix to create a Planary Maximum Filtered Graph
matrix inserting all the node and edge metrics into the graph
"""
if inv:
dist_matrix = 1.0 / df_distance.values
else:
dist_matrix = df_distance.values
index_upper_triangular = np.triu_indices(dist_matrix.shape[0], 1)
isort = np.argsort(dist_matrix[index_upper_triangular])
G = nx.Graph()
names = df_distance.columns.unique()
for k in np.arange(0, len(isort)):
u = index_upper_triangular[0][isort[k]]
v = index_upper_triangular[1][isort[k]]
if np.abs(
dist_matrix[u, v]
) > threshold / len(names): # optional bonferroni correction
edgedict = {'weight': float(dist_matrix[u, v])}
for item in matrix_panel.items:
edgedict[item] = float(matrix_panel.ix[item, u, v])
G.add_edge(u, v, edgedict)
if not planarity.is_planar(G):
G.remove_edge(u, v)
labels = {}
for i, n in enumerate(names):
name = names[i]
labels[i] = name
G = nx.relabel_nodes(G, labels)
return G
def construct_pmfg(self, df_distance, matrix_panel, inv=False, threshold=0.0):
"""Trim the provided adjacency matrix to create a Planary Maximum Filtered Graph
matrix inserting all the node and edge metrics into the graph
"""
if inv:
dist_matrix = 1.0/df_distance.values
else:
dist_matrix = df_distance.values
index_upper_triangular = np.triu_indices(dist_matrix.shape[0], 1)
isort = np.argsort(dist_matrix[index_upper_triangular])
G = nx.Graph()
names = df_distance.columns.unique()
for k in np.arange(0, len(isort)):
u = index_upper_triangular[0][isort[k]]
v = index_upper_triangular[1][isort[k]]
if np.abs(dist_matrix[u, v]) > threshold / len(names): # optional bonferroni correction
edgedict = {'weight': float(dist_matrix[u, v])}
for item in matrix_panel.items:
edgedict[item] = float(matrix_panel.ix[item, u, v])
G.add_edge(u, v, edgedict)
if not planarity.is_planar(G):
G.remove_edge(u, v)
labels = {}
for i, n in enumerate(names):
name = names[i]
labels[i] = name
G = nx.relabel_nodes(G, labels)
return G
def network_is_planar(network):
"""Check if the network is planar.
A network is planar if it can be drawn in the plane without crossing edges.
If a network is planar, it can be shown that an embedding of the network in
the plane exists, and, furthermore, that straight-line embedding in the plane
exists.
Warning:
This function uses the python binding of the *edge addition planarity suite*
in the background. The package is available on GitHub: https://github.com/hagberg/planarity.
Parameters:
network (compas.datastructures.network.Network): The network object.
Returns:
bool: ``True`` if the network is planar. ``False`` otherwise.
Raises:
ImportError: If the planarity package is not installed.
Example:
.. plot::
:include-source:
import compas
from compas.datastructures.network import Network
from compas.visualization.plotters import NetworkPlotter
from compas.datastructures.network.algorithms import network_is_planar
from compas.datastructures.network.algorithms import network_find_crossings
network = Network.from_obj(compas.get_data('lines.obj'))
network.add_edge(21, 29)
network.add_edge(17, 28)
if not network_is_planar(network):
crossings = network_find_crossings(network)
else:
crossings = []
plotter = NetworkPlotter(network)
plotter.draw_vertices(radius=0.15, text={key: key for key in network.vertices()})
plotter.draw_edges(color={edge: '#ff0000' for edges in crossings for edge in edges})
plotter.show()
"""
try:
import planarity
except ImportError:
print(
"Planarity is not installed. Get Planarity at https://github.com/hagberg/planarity."
)
raise
return planarity.is_planar(network.edges())
def graphFactoryPlanarErdosRenyiGenration(nb_graph,size_graph,graphtype,edgeProba,section='all'):
cpt=0
while cpt <= nb_graph:
G = nx.gnp_random_graph(size_graph,edgeProba)
if graphToCSV(G,graphtype,section,pl.is_planar(G)):
cpt+=1
if cpt%10 == 0:
print(str(cpt)+'/'+str(nb_graph)+' '+str(100*cpt/nb_graph)+'%')
def test_is_planar_unions(self):
try:
from itertools import combinations,product
except ImportError:
raise SkipTest('itertools.combinations not found')
for (G1,G2) in combinations(self.planar,2):
G=nx.disjoint_union(G1,G2)
assert_true(planarity.is_planar(G))
for (G1,G2) in combinations(self.non_planar,2):
G=nx.disjoint_union(G1,G2)
assert_false(planarity.is_planar(G))
for (G1,G2) in product(self.planar,self.non_planar):
G=nx.disjoint_union(G1,G2)
assert_false(planarity.is_planar(G))
def test_is_planar_unions(self):
try:
from itertools import combinations, product
except ImportError:
raise SkipTest('itertools.combinations not found')
for (G1, G2) in combinations(self.planar, 2):
G = nx.disjoint_union(G1, G2)
assert_true(planarity.is_planar(G))
for (G1, G2) in combinations(self.non_planar, 2):
G = nx.disjoint_union(G1, G2)
assert_false(planarity.is_planar(G))
for (G1, G2) in product(self.planar, self.non_planar):
G = nx.disjoint_union(G1, G2)
assert_false(planarity.is_planar(G))
def graphFactoryPlanar(nb_graph, size_graph, graphtype, section='all'):
cpt = 0
while cpt <= nb_graph:
m = np.random.random_sample(1)
G = nx.gnm_random_graph(size_graph,edgeForDensity(size_graph,m))
if graphToCSV(G,graphtype,section,pl.is_planar(G)):
cpt+=1
if cpt%10 == 0:
print(str(cpt)+'/'+str(nb_graph)+' '+str(100*cpt/nb_graph)+'%')
def graphFactoryPlanarNormalDistribution(nb_graph,size_graph,graphtype,location,spread,section='all'):
cpt=0
while cpt <= nb_graph:
rdm_density = np.random.normal(location,spread)
G = nx.gnm_random_graph(size_graph,edgeForDensity(size_graph,rdm_density))
if graphToCSV(G,graphtype,section,pl.is_planar(G)):
cpt+=1
if cpt%10 == 0:
print(str(cpt)+'/'+str(nb_graph)+' '+str(100*cpt/nb_graph)+'%')
def get_network_PMFG(corr_matrix):
"""
This function creates a filtered network starting from a correlation matrix using PMFG algorithm
:param corr_matrix (numpy 2D-array)
:return: returns filtered network (networkx Graph)
"""
# get the list of links in decreasing order of weight
rholist = []
n = len(corr_matrix)
for i in range(n):
for j in range(n):
if i > j: # matrix is symmetric
if corr_matrix[i][j] != 0:
rholist.append([abs(float(corr_matrix[i, j])), i,
j]) # weight, node1, node2
# sort the list in decreasing order
rholist.sort(key=lambda x: x[0])
rholist.reverse()
# initialize filtered matrix
m = len(rholist)
filtered_matrix = np.zeros((n, n))
control = 0
# filtered graph
G = nx.Graph()
# get the filtered adjacency matrix using the PMFG algorithm
# iterate over ordered edges
for t in range(m):
# stopping condition
if control <= 3 * (n - 2) - 1:
# get the current edge
i = rholist[t][1]
j = rholist[t][2]
filtered_matrix[i, j] = rholist[t][0]
# add the edge to the Graph
G.add_edge(int(i), int(j), weight=filtered_matrix[i, j])
# if the obtained G is not planarity we remove the edge
if planarity.is_planar(G) == False:
# remove edge
filtered_matrix[i, j] = 0
control = control + 1
G.remove_edge(int(i), int(j))
else:
break
return filtered_matrix, G
def test_goldner_harary(self):
# goldner-harary graph
# http://en.wikipedia.org/wiki/Goldner%E2%80%93Harary_graph
# a maximal planar graph
e= [(1,2 ),( 1,3 ),( 1,4 ),( 1,5 ),( 1,7 ),( 1,8 ),( 1,10 ),
( 1,11 ),( 2,3 ),( 2,4 ),( 2,6 ),( 2,7 ),( 2,9 ),( 2,10 ),
( 2,11 ),( 3,4 ),( 4,5 ),( 4,6 ),( 4,7 ),( 5,7 ),( 6,7 ),
( 7,8 ),( 7,9 ),( 7,10 ),( 8,10 ),( 9,10 ),( 10,11)]
G=nx.Graph(e)
assert_true(planarity.is_planar(G))
def test_goldner_harary(self):
# goldner-harary graph
# http://en.wikipedia.org/wiki/Goldner%E2%80%93Harary_graph
# a maximal planar graph
e = [(1, 2), (1, 3), (1, 4), (1, 5), (1, 7), (1, 8), (1, 10), (1, 11),
(2, 3), (2, 4), (2, 6), (2, 7), (2, 9), (2, 10), (2, 11), (3, 4),
(4, 5), (4, 6), (4, 7), (5, 7), (6, 7), (7, 8), (7, 9), (7, 10),
(8, 10), (9, 10), (10, 11)]
G = nx.Graph(e)
assert_true(planarity.is_planar(G))
def compute_PMFG(sorted_edges, nb_nodes):
PMFG = nx.Graph()
for edge in sorted_edges:
PMFG.add_edge(edge[0], edge[1], metric=edge[2])
if not planarity.is_planar(PMFG):
PMFG.remove_edge(edge[0], edge[1])
if len(PMFG.edges()) == 3 * (nb_nodes - 2):
break
return PMFG
def compute_PMFG(sorted_edges, nb_nodes):
PMFG = nx.Graph()
for edge in sorted_edges:
PMFG.add_edge(edge['source'], edge['dest'])
if not planarity.is_planar(PMFG):
PMFG.remove_edge(edge['source'], edge['dest'])
if len(PMFG.edges()) == 3 * (nb_nodes - 2):
break
return PMFG
def is_planar(self):
"""Verify that the graph has a planar embedding.
Returns
-------
bool
"""
if compas.IPY:
from compas.rpc import Proxy
planarity = Proxy('planarity')
else:
import planarity
return planarity.is_planar(list(self.edges()))
def construct_pmfg(df_corr_matrix):
df_distance = compute_distance(df_corr_matrix)
dist_matrix = df_distance.values
index_upper_triangular = np.triu_indices(dist_matrix.shape[0], 1)
isort = np.argsort(dist_matrix[index_upper_triangular])
G = nx.Graph()
for k in range(len(isort)):
u = index_upper_triangular[0][isort[k]]
v = index_upper_triangular[1][isort[k]]
if dist_matrix[u, v] > .0: # remove perfect correlation because of diagonal FIXME
G.add_edge(u, v, {'weight': float(dist_matrix[u, v])})
if not planarity.is_planar(G):
G.remove_edge(u, v)
return G
def construct_pmfg(df_corr_matrix):
# TW: Is this required? What's the difference to graph.algos.construct_pmfg?
df_distance = np.sqrt( 2 * np.clip( 1. - df_corr_matrix, 0., 2.) )
dist_matrix = df_distance.values
index_upper_triangular = np.triu_indices(dist_matrix.shape[0],1)
isort = np.argsort( dist_matrix[index_upper_triangular] )
G = nx.Graph()
for k in xrange(0,len(isort)):
u = index_upper_triangular[0][isort[k]]
v = index_upper_triangular[1][isort[k]]
if dist_matrix[u,v] > 0: # remove perfect correlation because of diagonal FIXME
G.add_edge(u, v, {'weight': float(dist_matrix[u,v])})
if not planarity.is_planar(G):
G.remove_edge(u,v)
return G
def brute_force_thickness(g):
smallest_thickness = 500
for thickness_guess in range(2, 4):
for edge_arrangement in algorithm_u(g.edges(), thickness_guess):
all_planar = True
for layer in edge_arrangement:
if not is_planar(_from_edge_list(layer)):
all_planar = False
break
if all_planar and len(edge_arrangement) < smallest_thickness:
smallest_thickness = len(edge_arrangement)
break
if smallest_thickness <= thickness_guess:
break
return smallest_thickness
def check_planar(dataset):
with open("generated_smiles_%s" % dataset, 'rb') as f:
all_smiles = set(pickle.load(f))
total_non_planar = 0
for smiles in all_smiles:
try:
nodes, edges = to_graph(smiles, dataset)
except:
continue
edges = [(src, dst) for src, e, dst in edges]
if edges == []:
continue
if not planarity.is_planar(edges):
total_non_planar += 1
return len(all_smiles), total_non_planar
def simplify_using_PMFG(corr_matrix):
#get the list of decreasing weighted links
rholist = []
n = len(corr_matrix)
for i in range(n):
for j in range(n):
if i < j:
if corr_matrix[i][j] != 0:
rholist.append([abs(float(corr_matrix[i][j])), i, j])
rholist.sort(key=lambda x: x[0])
rholist.reverse()
m = len(rholist)
filtered_matr = np.zeros((n, n))
control = 0
with progressbar.ProgressBar(max_value=m) as bar:
#get the filtered adjacency matrix using PMFG algorithm
for t in range(m):
if control <= 3 * (n - 2) - 1:
i = rholist[t][1]
j = rholist[t][2]
filtered_matr[i][j] = rholist[t][0]
#check planarity here
G = nx.Graph()
for i in range(0, n):
for j in range(0, n):
if filtered_matr[i][j] != 0:
G.add_edge(int(i),
int(j),
weight=filtered_matr[i][j])
if planarity.is_planar(G) == False:
filtered_matr[i][j] = 0
control = control + 1
bar.update(t)
#build the network
PMFG = nx.Graph()
for i in range(0, n):
for j in range(0, n):
if filtered_matr[i][j] != 0:
PMFG.add_edge(int(i), int(j), weight=filtered_matr[i][j])
return PMFG
def construct_pmfg(df_corr_matrix):
# TW: Is this required? What's the difference to graph.algos.construct_pmfg?
df_distance = np.sqrt(2 * np.clip(1. - df_corr_matrix, 0., 2.))
dist_matrix = df_distance.values
index_upper_triangular = np.triu_indices(dist_matrix.shape[0], 1)
isort = np.argsort(dist_matrix[index_upper_triangular])
G = nx.Graph()
for k in xrange(0, len(isort)):
u = index_upper_triangular[0][isort[k]]
v = index_upper_triangular[1][isort[k]]
if dist_matrix[
u,
v] > 0: # remove perfect correlation because of diagonal FIXME
G.add_edge(u, v, {'weight': float(dist_matrix[u, v])})
if not planarity.is_planar(G):
G.remove_edge(u, v)
return G
def is_planar(location, adj_list, is_dense=False):
if is_dense:
new_adj_list = defaultdict(list)
for x in range(len(adj_list)):
for y in range(len(adj_list)):
if adj_list[x][y] == 1:
new_adj_list[x].append((y, 1))
adj_list = new_adj_list
edges = []
seen = set()
for src, l in adj_list.items():
for dst, e in l:
if (dst, src) not in seen:
edges.append((src, dst))
seen.add((src, dst))
edges += [location, (location[1], location[0])]
return planarity.is_planar(edges)
def network_is_planar(network):
"""Check if the network is planar.
Parameters
----------
network : Network
A network object.
Returns
-------
bool
True if the network is planar.
False otherwise.
Raises
------
ImportError
If the planarity package is not installed.
Notes
-----
A network is planar if it can be drawn in the plane without crossing edges.
If a network is planar, it can be shown that an embedding of the network in
the plane exists, and, furthermore, that straight-line embedding in the plane
exists.
Warnings
--------
This function uses the python binding of the *edge addition planarity suite*.
It is available on Anaconda: https://anaconda.org/conda-forge/python-planarity.
Examples
--------
>>>
"""
try:
import planarity
except ImportError:
print("Planarity is not installed.")
raise
return planarity.is_planar(list(network.edges()))
def compute_PMFG(sorted_edges, nb_nodes):
# remove weakest links
'''
References
Tumminello, M., Aste, T., Di Matteo, T., & Mantegna, R. N. (2005). A tool for filtering information in complex systems. Proceedings of the National Academy of Sciences of the United States of America, 102(30), 10421-10426.
https://gmarti.gitlab.io/networks/2018/06/03/pmfg-algorithm.html
https://github.com/hagberg/planarity
http://jgaa.info/accepted/2004/BoyerMyrvold2004.8.3.pdf
https://networkx.github.io/documentation/stable/index.html
'''
PMFG = nx.Graph()
for edge in sorted_edges:
PMFG.add_edge(edge['source'], edge['dest'], weight=edge['weight'])
if not planarity.is_planar(PMFG):
PMFG.remove_edge(edge['source'], edge['dest'])
if len(PMFG.edges()) == 3 * (nb_nodes - 2):
break
return PMFG
def compute_PMFG(G):
PMFG = nx.Graph() # initialize
ne_total = G.number_of_edges()
nb_nodes = len(G.nodes)
ne_pmfg = 3 * (nb_nodes - 2)
sorted_edges = sort_graph_edges_corr(G)
t0 = time.time()
for i, edge in enumerate(sorted_edges):
PMFG.add_edge(edge['source'], edge['dest'], weight=edge['weight'])
if not planarity.is_planar(PMFG):
PMFG.remove_edge(edge['source'], edge['dest'])
ne = PMFG.number_of_edges()
print(
"Generating PMFG... added edges in PMFG %d/%d (%.2f%%) lookup edges in G %d/%d (%.2f%%) Elapsed TIme %.2f [sec]"
% (ne, ne_pmfg, (ne / ne_pmfg) * 100, i, ne_total,
(i + 1 / ne_total) * 100, time.time() - t0),
end="\r")
if ne == ne_pmfg:
break
return PMFG
def exportRandomGraph(name, path=''):
G = nx.newman_watts_strogatz_graph(6,2,0.5)
nx.draw(G)
plt.savefig('img_rdm_generate_graph_'+name+'.png')
plt.clf()
print(pl.is_planar(G))
def test_is_planar(self):
for G in self.planar:
assert_true(planarity.is_planar(G))
for G in self.non_planar:
assert_false(planarity.is_planar(G))
import planarity import networkx as nx # Example of the complete graph of 5 nodes, K5 G=nx.complete_graph(5) # K5 is not planar print(planarity.is_planar(G)) # False # find forbidden Kuratowski subgraph K=planarity.kuratowski_subgraph(G) print(K.edges()) # K5 edges
def is_planar(edge_list):
return planarity.is_planar(edge_list)
import planarity
n = 6
m = n * (n - 1) / 2
out = "1"
for x in xrange(1, pow(2, m)):
edgelist = []
u, v = 1, 1
for y in xrange(m):
u += 1
if u >= v:
v += 1
u = 1
if x & (1<<(m-y-1)):
edgelist.append((u, v))
out += "1" if planarity.is_planar(edgelist) else "0"
print(out)
import planarity
# Example of the complete graph of 5 nodes, K5
# K5 is not planar
# use text strings as labels
edgelist = [('a', 'b'), ('a', 'c'), ('a', 'd'), ('a', 'e'),
('b', 'c'),('b', 'd'),('b', 'e'),
('c', 'd'), ('c', 'e'),
('d', 'e')]
print planarity.is_planar(edgelist) # False
# print forbidden Kuratowski subgraph (K5)
print planarity.kuratowski_edges(edgelist)
# remove an edge
edgelist.remove(('a','b'))
# graph is now planar
print planarity.is_planar(edgelist) # True
# no forbidden subgraph, empty list returned
print planarity.kuratowski_edges(edgelist)
def test_is_planar(self):
for G in self.planar:
assert_true(planarity.is_planar(G))
for G in self.non_planar:
assert_false(planarity.is_planar(G))
import planarity
edgelist = [
('a', 'b'), ('a', 'c'), ('a', 'd'), ('a', 'e'),
('b', 'c'),('b', 'd'),('b', 'e'),
('c', 'd'), ('c', 'e'),
('d', 'e')
]
print(planarity.is_planar(edgelist))
nx.write_adjlist(new_G, output_path_adj)
if write_graph:
if verbose:
print('Saving graph: %s' % output_path)
sys.stdout.flush()
graphutils.write_graph(new_G, output_path)
image_path = output_path + '.pdf'
stderr_path = output_path + '.err.txt'
if init_options['compare_replica']:
if verbose:
print('Generator Report')
print('Comparing replica')
sys.stdout.flush()
graphutils.compare_nets(G, new_G, params=params)
print(planarity.is_planar(new_G.edges()))
pos = nx.graphviz_layout(new_G)
nx.draw(new_G, pos, with_labels=False, node_size=1)
plt.show()
benchmarks.find_differences(G, new_G)
#0.03 is too small for Linux
#sfdp_default_cmd = 'sfdp -Goverlap="prism100" -Goverlap_scaling=-100 -Nlabel="" -Nwidth=0.01 -Nfixedsize=true -Nheight=0.01'
sfdp_default_cmd = 'sfdp -Nlabel="" -Nwidth=0.06 -Nfixedsize=true -Nheight=0.06 -Nstyle=filled'
if write_graph and visualizer == 'sfdp' and output_path[-3:] == 'dot':
visualizer_cmdl = sfdp_default_cmd + ' -Tpdf %s > %s 2> %s ' % (
output_path, image_path, stderr_path)
if verbose:
print('Writing graph image: %s ..' % image_path)
sys.stdout.flush()
retCode = os.system(visualizer_cmdl)
if __name__ == '__main__':
arguments = docopt(__doc__)
M = int(arguments['M'])
N = int(arguments['N'])
output = arguments['OUTPUT']
# G = nx.complete_multipartite_graph(M, N)
# G = nx.complete_graph(M)
G = nx.gnm_random_graph(M, N)
if planarity_test.is_planar(G):
print("Graph is planar")
try:
draw(G, output)
print("Planar graph drawn!")
except:
print("Could not draw a planar graph...")
nx.draw_random(G)
plt.axis('off')
plt.savefig(output)
else:
if planarity.is_planar(G):
draw(G, output)
print("Graph is unexpectedly planar...")
else:
nx.draw_random(G)
plt.axis('off')
plt.savefig(output)
print("Graph is not planar!")
def exportGraph(G, name, path=''):
nx.draw(G)
plt.savefig('img_'+name+'.png')
plt.clf()
print(pl.is_planar(G))
def network_is_planar(network):
"""Check if the network is planar.
Parameters
----------
network : Network
A network object.
Returns
-------
bool
True if the network is planar.
False otherwise.
Raises
------
ImportError
If the planarity package is not installed.
Notes
-----
A network is planar if it can be drawn in the plane without crossing edges.
If a network is planar, it can be shown that an embedding of the network in
the plane exists, and, furthermore, that straight-line embedding in the plane
exists.
Warning
-------
This function uses the python binding of the *edge addition planarity suite*.
It is available on GitHub: https://github.com/hagberg/planarity.
Examples
--------
.. plot::
:include-source:
import compas
from compas.datastructures import Network
from compas.topology import network_is_planar
from compas.topology import network_find_crossings
from compas.plotters import NetworkPlotter
network = Network.from_obj(compas.get('lines.obj'))
network.add_edge(21, 29)
network.add_edge(17, 28)
if not network_is_planar(network):
crossings = network_find_crossings(network)
else:
crossings = []
plotter = NetworkPlotter(network)
plotter.draw_vertices(radius=0.15, text={key: key for key in network.vertices()})
plotter.draw_edges(color={edge: '#ff0000' for edges in crossings for edge in edges})
plotter.show()
"""
try:
import planarity
except ImportError:
print(
"Planarity is not installed. Get Planarity at https://github.com/hagberg/planarity."
)
raise
return planarity.is_planar(network.edges())
# K5 is not planar
# any of the following formats can bed used for representing the graph
edgelist = [(0, 1), (0, 2), (0, 3), (0, 4),
(1, 2),(1, 3),(1, 4),
(2, 3), (2, 4),
(3, 4)]
dictofdicts = {0: {1: {}, 2: {}, 3: {}, 4: {}},
1: {2: {}, 3: {}, 4: {}},
2: {3: {}, 4: {}},
3: {4: {}},
4: {}}
dictofsets = {0: set([1,2,3,4]),
1: set([2,3,4]),
2: set([3,4]),
3: set([4]),
4: set([])}
dictoflists = {0: list([1,2,3,4]),
1: list([2,3,4]),
2: list([3,4]),
3: list([4]),
4: list([])}
print planarity.is_planar(edgelist) # False
print planarity.is_planar(dictofdicts) # False
print planarity.is_planar(dictofsets) # False
print planarity.is_planar(dictoflists) # False
def test_is_planar_adj_input_function(self):
assert_false(planarity.is_planar(self.k5_adj))
def network_is_planar(network):
"""Check if the network is planar.
Parameters
----------
network : Network
A network object.
Returns
-------
bool
True if the network is planar.
False otherwise.
Raises
------
ImportError
If the planarity package is not installed.
Notes
-----
A network is planar if it can be drawn in the plane without crossing edges.
If a network is planar, it can be shown that an embedding of the network in
the plane exists, and, furthermore, that straight-line embedding in the plane
exists.
Warning
-------
This function uses the python binding of the *edge addition planarity suite*.
It is available on Anaconda: https://anaconda.org/conda-forge/python-planarity.
Examples
--------
.. plot::
:include-source:
import compas
from compas.datastructures import Network
from compas.datastructures import network_is_planar
from compas.datastructures import network_find_crossings
from compas_plotters import NetworkPlotter
network = Network.from_obj(compas.get('lines.obj'))
network.add_edge(6, 15)
if not network_is_planar(network):
print('here')
crossings = network_find_crossings(network)
else:
crossings = []
print(crossings)
plotter = NetworkPlotter(network)
plotter.draw_vertices(radius=0.15, text={key: key for key in network.vertices()})
plotter.draw_edges(color={edge: '#ff0000' for edges in crossings for edge in edges})
plotter.show()
"""
return planarity.is_planar(list(network.edges()))
import planarity import networkx as nx G8 = nx.complete_graph(8) G8.nodes G8.edges planarity.is_planar(G8) K = planarity.kuratowski_subgraph(G8) K.edges K.nodes nx.draw(G8)
def test_is_planar_edgelist_input_function(self):
assert_false(planarity.is_planar(self.k5_edgelist))