G.node[nodeid]['gender'] = gender_code[0]
return G
G = load_crime_network()
# Annotate each node with connectivity score
for n in G.nodes():
dcs = nx.degree_centrality(G)
G.node[n]['connectivity'] = dcs[n]
# Make a CircosPlot of the bipartite graph
c = CircosPlot(
G,
node_grouping='bipartite',
node_order='connectivity',
node_color='bipartite',
)
c.draw()
# Make the "people" projection of the bipartite graph.
person_nodes = [n for n in G.nodes() if G.node[n]['bipartite'] == 'person']
pG = nx.bipartite.projection.projected_graph(G, person_nodes)
for n in pG.nodes():
dcs = nx.degree_centrality(pG)
pG.node[n]['connectivity'] = dcs[n]
c = CircosPlot(pG,
node_grouping='gender',
node_order='connectivity',
def createGraph():
# read data from database
conn = sqlite3.connect('temp.db')
cursor = conn.cursor()
df_author = []
sql_jn = """
SELECT journal_name
FROM journal_list
"""
for row in cursor.execute(sql_jn):
jn = str(row[0])
print(jn)
sql_final = """
SELECT {paper_country_temp}.paper_id, country_list.country_name
FROM {paper_country_temp}
INNER JOIN country_list ON country_list.country_id = {paper_country_temp}.country_id
;""".format(paper_country_temp='paper_country_' +
jn.replace(" ", "_"))
df2 = pd.read_sql_query(sql_final, conn)
gb = df2['country_name'].groupby([df2.paper_id
]).apply(list).reset_index()
df_author.append(gb['country_name'].tolist())
print('---------------')
flatten = [item for elem in df_author for item in elem]
conn.commit()
conn.close()
record_list = []
# part 2 $$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
dict_country = {
'the Netherlands': 'Netherlands',
'The Netherlands': 'Netherlands',
'Aero Engine (Group) Corporation of China': 'China',
'PR China': 'China',
'United States': 'USA',
'United Kingdom': 'UK',
"People's Republic of China": 'China',
'Republic of Singapore': 'Singapore',
'United States of America': 'USA',
'P.R. China': 'China',
}
sns.set_style("white")
authors_flat = [author for authors in flatten for author in authors]
# Part 3 $$$$$$$$$$$$$$$$$$$$$$$$$$$
# Extract author connections article_authors
authors = flatten
print('@@@@@@@@@@@@@@@@@@@@@@@@@@')
print(type(authors))
print('%%%%%%%%%%%%%%%%%%%%%%%%%%')
author_connections = list(
map(lambda x: list(combinations(x[::-1], 2)), authors))
flat_connections = [
item for sublist in author_connections for item in sublist
]
print(len(flat_connections))
# Create a dataframe with the connections
df = pd.DataFrame(flat_connections, columns=["From", "To"])
df['From'] = df['From'].replace(dict_country)
df['To'] = df['To'].replace(dict_country)
df = df[df['From'] != df['To']]
print(df)
# df.to_excel('1.xlsx')
df_graph = df.groupby(["From", "To"]).size().reset_index()
print(df_graph)
# df_graph.to_excel('2.xlsx')
df_graph.columns = ["From", "To", "Count"]
print(df_graph)
G = nx.from_pandas_edgelist(df_graph,
source="From",
target="To",
edge_attr="Count")
# Limit to TOP 50 authors
top50authors = pd.DataFrame.from_records(
Counter(authors_flat).most_common(50), columns=["Name", "Count"])
top50_nodes = (n for n in list(G.nodes())
if n in list(top50authors["Name"]))
G_50 = G.subgraph(top50_nodes)
for n in G_50.nodes():
G_50.node[n]["publications"] = int(
top50authors[top50authors["Name"] == n]["Count"])
c = CircosPlot(
G_50,
dpi=600,
node_grouping="publications",
figsize=(20, 20),
node_color="publications",
node_labels=True,
)
c.draw()
plt.show()
# find path
# paths = list(
# nx.all_shortest_paths(G, source="Xing Wang", target="Wen Li")
# )
# for path in paths:
# print(path)
return c
G.node[nodeid]["gender"] = gender_code[0]
return G
G = load_crime_network()
# Annotate each node with connectivity score
for n in G.nodes():
dcs = nx.degree_centrality(G)
G.node[n]["connectivity"] = dcs[n]
# Make a CircosPlot of the bipartite graph
c = CircosPlot(
G,
node_grouping="bipartite",
node_order="connectivity",
node_color="bipartite",
)
c.draw()
# Make the "people" projection of the bipartite graph.
person_nodes = [n for n in G.nodes() if G.node[n]["bipartite"] == "person"]
pG = nx.bipartite.projection.projected_graph(G, person_nodes)
for n in pG.nodes():
dcs = nx.degree_centrality(pG)
pG.node[n]["connectivity"] = dcs[n]
c = CircosPlot(pG,
node_grouping="gender",
node_order="connectivity",
edge_list=[]
for i in range(aa11.shape[0]):
node=i+1
node_list.append(node)
for i in range(len(last_coor)):
last_coor[i][0] += 1
last_coor[i][1] += 1
edge_list.append(last_coor[i])
G = nx.Graph()
G.add_nodes_from(node_list)
G.add_edges_from(edge_list)
color_list=["a", "b", "c", "d", "e"]
for n, d in G.nodes(data=True):
G.node[n]["class"] = node_list[n-1]
c = CircosPlot(graph=G,node_labels=True,
node_label_rotation=True,
fontsize=30,
group_legend=True,
figsize=(7, 7),node_color="class")
c.draw()
plt.title("circular graph of top Nf discriminative features across all cross-validation runs".title())
plt.show()
#* *
from random import choice
import matplotlib.pyplot as plt
import networkx as nx
from nxviz.plots import CircosPlot
G = nx.barbell_graph(m1=10, m2=3)
for n, d in G.nodes(data=True):
G.node[n]['class'] = choice(['a', 'b', 'c', 'd', 'e'])
c = CircosPlot(G,
node_grouping="class",
node_color="class",
node_order='class',
node_labels=False,
group_label_position="middle",
group_label_color=True)
c.draw()
plt.show()
def create_circos(data,
source,
target,
edge_attr,
node_classes,
node_grouping,
edge_width,
edge_color,
legend_label,
legend_color,
legend_sort,
node2color,
figsize=(15, 15),
alpha=.5,
title=None,
group_label_offset=10,
legend_length=9,
**kwargs):
"""Create a circos plot with correlations"""
# create nodes and edges
G = nx.from_pandas_edgelist(data,
source=source,
target=target,
edge_attr=edge_attr)
# classify a node
for n, _ in G.nodes(data=True):
G.nodes[n]["class"] = node_classes[n]
# initialize Circos plot
c = CircosPlot(G,
node_grouping=node_grouping,
edge_width=edge_width,
edge_color=edge_color,
figsize=figsize,
node_labels=True,
node_label_layout="rotation",
group_label_position=None,
group_label_color=True,
group_label_offset=group_label_offset,
**kwargs)
# set alpha of edges
c.edgeprops['alpha'] = alpha
# set node colors
node_colors = []
for l in c.nodes:
lc = l.split('(')[0].strip()
co = node2color[lc]
c_rgba = to_rgba(co)
node_colors.append(c_rgba)
c.node_colors = node_colors
c.node_size = .5
# set edge colors
edge_colors = []
for e in c.edges:
try:
ec = data.loc[(data[source] == e[0]) & (data[target] == e[1]),
legend_color]
edge_color = ec.item()
except ValueError:
ec = data.loc[(data[source] == e[1]) & (data[target] == e[0]),
legend_color]
edge_color = ec.item()
except Exception as err:
print(e)
print(ec)
raise err
edge_colors.append(edge_color)
c.edge_colors = edge_colors
# actually draw the Circos plot
c.draw()
# get matplotlib.pyplot figure to add legend and title
fig = c.figure
# create legend for edge colors
edge_legends, edge_labels = custom_legend(data=data,
title='Edge colors',
length=legend_length,
label=legend_label,
color=legend_color,
sort_by=legend_sort,
alpha=alpha)
# create legend for node colors
class2color = {}
for k, v in node_classes.items():
if k.startswith('UR_'): k = k.split('(')[0].strip()
class2color[v] = node2color[k]
node_legends, node_labels = custom_legend(data=class2color,
length=legend_length,
title='Node colors',
alpha=alpha)
# add legend
fig.legend(handles=edge_legends + node_legends,
labels=edge_labels + node_labels,
ncol=2,
loc='lower right')
if title is not None:
fig.suptitle(title, fontsize=16)
plt.close(fig)
return fig
""" Displays a NetworkX octahedral graph to screen using a CircosPlot. """ from nxviz.plots import CircosPlot import networkx as nx import matplotlib.pyplot as plt G = nx.octahedral_graph() c = CircosPlot(G.nodes(), G.edges(), plot_radius=5) c.draw() plt.show()
G = nx.Graph()
G.add_edges_from(zip(df['doctor1'], df['doctor2']))
return G
G = load_physicians_network()
# Make a CircosPlot, but with the nodes colored by their connected component
# subgraph ID.
ccs = nx.connected_component_subgraphs(G)
for i, g in enumerate(ccs):
for n in g.nodes():
G.node[n]['group'] = i
G.node[n]['connectivity'] = G.degree(n)
m = CircosPlot(G,
node_color='group',
node_grouping='group',
node_order='connectivity')
m.draw()
plt.show()
# Make an ArcPlot.
a = ArcPlot(G,
node_color='group',
node_grouping='group',
node_order='connectivity')
a.draw()
plt.show()
"""
Displays a NetworkX lollipop graph to screen using a CircosPlot.
"""
from nxviz.plots import CircosPlot
import numpy.random as npr
import networkx as nx
import matplotlib.pyplot as plt
G = nx.lollipop_graph(m=10, n=4)
for n, d in G.nodes(data=True):
G.node[n]['value'] = npr.normal()
c = CircosPlot(G, node_color='value', node_order='value')
c.draw()
plt.show()
plt.figure(figsize=(10, 7), dpi=150)
if type_plot == "classic":
# Classic graphviz plot
nx.draw(G,
with_labels=True,
node_shape="o",
node_color=node_color,
node_size=800,
font_size=8)
elif type_plot == "nxviz":
from nxviz.plots import CircosPlot as Plot
c = Plot(
G,
node_color="value",
node_order="value",
node_labels=True,
# , node_color=node_color
)
c.draw()
plt.tight_layout()
plt.savefig("graph.png")
plt.show()
try:
import pygraphviz
from networkx.drawing.nx_agraph import write_dot
print("using package pygraphviz")
except ImportError:
try:
import pydot
G = nx.Graph()
G.add_edges_from(zip(df["doctor1"], df["doctor2"]))
return G
G = load_physicians_network()
# Make a CircosPlot, but with the nodes colored by their connected component
# subgraph ID.
ccs = nx.connected_component_subgraphs(G)
for i, g in enumerate(ccs):
for n in g.nodes():
G.node[n]["group"] = i
G.node[n]["connectivity"] = G.degree(n)
m = CircosPlot(G,
node_color="group",
node_grouping="group",
node_order="connectivity")
m.draw()
plt.show()
# Make an ArcPlot.
a = ArcPlot(G,
node_color="group",
node_grouping="group",
node_order="connectivity")
a.draw()
plt.show()
continue
recognize_counts[recognizing_tf] += 1
G.add_node(target_tf, bipartite="target_tf", tf_fam=target_fam)
G.add_node(recognizing_tf, bipartite="recognizing_tf", tf_fam=recognizing_fam)
if target_tf != recognizing_tf:
G.add_edge(target_tf, recognizing_tf, target_fam=target_fam, recognizing_fam=recognizing_fam)
return G, recognize_counts
G,recognize_counts = load_motifs_network()
m = CircosPlot(
G, node_grouping="tf_fam", node_color="tf_fam",
group_label_position='middle', group_label_color=True,
#, node_order="tf_fam"
edge_color='target_fam',
node_labels= lambda node: recognize_counts[node] >= 4,
node_label_layout='rotation',
fontsize=8,
)
m.compute_node_colors()
m.draw()
plt.show()
# # # Make an ArcPlot.
# a = ArcPlot(
# G, node_color="tf_fam", node_grouping="tf_fam"#, node_order="connectivity"
# )
# a.draw()
# plt.show()
min_col += 1
g = nx.Graph(adj_matrix)
g = nx.relabel_nodes(g, nodes_idxs_rev)
for node in g.nodes():
if node in nodes_g1:
g.node[node]["class"] = "g1"
if node in nodes_g2:
g.node[node]["class"] = "g2"
if node in nodes_g3:
g.node[node]["class"] = "g3"
g = nx.relabel_nodes(g, term_names)
#c = CircosPlot(g, node_label_layout="numbers", dpi=800, fontsize=10,
# node_color="class", node_grouping="class",
# fig_size=(20,20), edge_width=weights, node_labels=True)
c = CircosPlot(g,
dpi=800,
fontsize=8,
node_label_layout="numbers",
node_color="class",
node_grouping="class",
fig_size=(20, 20),
edge_width=weights,
node_labels=True)
c.draw()
#plt.show()
plt.savefig("/home/wkg/Desktop/circos_test5.png", bbox_inches="tight", dpi=600)
),
"purple": ("Fortran", "P3DFFT"),
"slategrey": ("CUDA", "cuFFT", "OpenMP", "MPI", "SIMD"),
}
G = read_adjlist(root / "dependency/graph.txt")
node_color = []
for node in G.nodes():
color = [k for k, v in colors.items() if node in v]
if color:
# node_color.append(color[0])
G.node[node]["value"] = color[0]
else:
raise ValueError("No color assigned for %s" % node)
# nx.draw(G, with_labels=True, node_shape="s", node_color=node_color,
# node_size=1000,)
from nxviz.plots import CircosPlot
c = CircosPlot(
G,
node_color="value",
node_order="value",
node_labels=True,
# , node_color=node_color
fontfamily="monospace",
)
c.draw()
plt.tight_layout()
plt.savefig(root / "dependency.pdf")
# plt.show()
def createGraphAuthor(user_year, input1):
# read data from database
conn = sqlite3.connect('temp.db')
cursor = conn.cursor()
year_list = list(range(user_year[0], user_year[1] + 1, 1))
if len(year_list) == 1:
year_list.append(year_list[0])
df_author = []
sql_jn = """
SELECT journal_name
FROM journal_list
"""
for row in cursor.execute(sql_jn):
jn = str(row[0])
print(jn)
sql_final = """
SELECT {paper_author_temp}.paper_id, author_list.author_name
FROM {paper_author_temp}
INNER JOIN author_list ON author_list.author_id = {paper_author_temp}.author_id
WHERE paper_id IN (SELECT paper_id
FROM {temp_paper_list}
WHERE paper_abstract LIKE '%{user_search_temp}%'
)
AND
paper_id IN (SELECT paper_id
FROM {temp_paper_list}
WHERE paper_year IN ( SELECT year_id
FROM year_list
WHERE year IN {temp_year}
))
;""".format(paper_author_temp='paper_author_' + jn.replace(" ", "_"),
temp_paper_list='paper_list_' + jn.replace(" ", "_"),
user_search_temp=input1,
temp_year=str(tuple(year_list)))
df2 = pd.read_sql_query(sql_final, conn)
gb = df2['author_name'].groupby([df2.paper_id
]).apply(list).reset_index()
df_author.append(gb['author_name'].tolist())
print('---------------')
flatten = [item for elem in df_author for item in elem]
conn.commit()
conn.close()
record_list = []
# part 2 $$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
sns.set_style("white")
authors_flat = [author for authors in flatten for author in authors]
# Part 3 $$$$$$$$$$$$$$$$$$$$$$$$$$$
# Extract author connections article_authors
authors = flatten
print('@@@@@@@@@@@@@@@@@@@@@@@@@@')
print(type(authors))
print('%%%%%%%%%%%%%%%%%%%%%%%%%%')
author_connections = list(
map(lambda x: list(combinations(x[::-1], 2)), authors))
flat_connections = [
item for sublist in author_connections for item in sublist
]
print(len(flat_connections))
# Create a dataframe with the connections
df = pd.DataFrame(flat_connections, columns=["From", "To"])
print(df)
# df.to_excel('1.xlsx')
df_graph = df.groupby(["From", "To"]).size().reset_index()
print(df_graph)
# df_graph.to_excel('2.xlsx')
df_graph.columns = ["From", "To", "Count"]
print(df_graph)
G = nx.from_pandas_edgelist(df_graph,
source="From",
target="To",
edge_attr="Count")
# Limit to TOP 50 authors
top50authors = pd.DataFrame.from_records(
Counter(authors_flat).most_common(50), columns=["Name", "Count"])
top50_nodes = (n for n in list(G.nodes())
if n in list(top50authors["Name"]))
G_50 = G.subgraph(top50_nodes)
for n in G_50.nodes():
G_50.node[n]["publications"] = int(
top50authors[top50authors["Name"] == n]["Count"])
c = CircosPlot(
G_50,
dpi=600,
node_grouping="publications",
edge_width="Count",
figsize=(20, 20),
node_color="publications",
node_labels=True,
)
m = c.draw()
return m
import matplotlib.pyplot as plt
import networkx as nx
from nxviz.plots import CircosPlot
G = nx.barbell_graph(m1=10, m2=3)
for n, d in G.nodes(data=True):
G.node[n]["class"] = choice(["a", "b", "c", "d", "e"])
c = CircosPlot(G,
node_grouping="class",
node_color="class",
node_order="class",
node_labels=True,
node_label_rotation=True,
node_label_color=True,
group_label_color=True,
node_label_layout='numbers',
group_order='alphabetically',
fontsize=10,
group_legend=True,
figsize=(12, 12))
# Draw the CircosPlot
c.draw()
c.figure.tight_layout()
# If you want more control over the legend, you can pass c.legend_handles
# to plt.legend()
plt.legend(handles=c.legend_handles,
loc="upper center",
"""
Shows how to rotate node labels. This increaes legibility for longer labels.
"""
import matplotlib.pyplot as plt
import networkx as nx
from nxviz.plots import CircosPlot
G = nx.barbell_graph(m1=20, m2=3)
# let's give the nodes some longer labels
G = nx.relabel_nodes(G,
{i: "long name #" + str(i) for i in range(len(G))}
)
# try it `rotate_labels=False` to see how the long names overlap each other
c = CircosPlot(G, node_labels=True, rotate_labels=True)
c.draw()
# the rotated labels take up more space, so we will have to increase the
# padding a bit. 15% on all sides works well here.
plt.tight_layout(rect=(0.15, 0.15, 0.85, 0.85))
plt.show()
""" Displays a NetworkX octahedral graph to screen using a CircosPlot. """ import matplotlib.pyplot as plt import networkx as nx from nxviz.plots import CircosPlot G = nx.octahedral_graph() c = CircosPlot(G) c.draw() plt.show()
"""
Displays a NetworkX barbell graph to screen using a CircosPlot.
"""
from random import choice
import matplotlib.pyplot as plt
import networkx as nx
from nxviz.plots import CircosPlot
G = nx.barbell_graph(m1=10, m2=3)
for n, d in G.nodes(data=True):
G.node[n]['class'] = choice(['one', 'two', 'three'])
c = CircosPlot(G, node_color="class", node_order='class', node_labels=True)
c.draw()
plt.show()
pos = nx.spring_layout(G) # positions for all nodes
# # First, draw nodes from table load
nx.draw_networkx_nodes(G, pos, nodelist=views, node_color='b')
nx.draw_networkx_nodes(G, pos, nodelist=tables, node_color='r')
nx.draw_networkx_edges(G, pos, edgelist=tl_edges, edge_color='b')
nx.draw_networkx_edges(G, pos, edgelist=vd_edges, edge_color='r')
nx.draw_networkx_labels(G, pos)
nx.draw_networkx_edge_labels(G, pos)
# nx.draw_networkx_labels(G)
plt.show()
# some plots
#cpt = CircosPlot(graph=G, node_labels=True, node_color='label', edge_color='label')
#
cpt = CircosPlot(graph=G, node_labels=True, node_color='schema')
cpt.draw()
plt.show()
nx.draw_kamada_kawai(G)
# nx.draw_networkx_labels(G)
# nx.draw_networkx_edge_labels(G)
plt.show()
# plt.savefig("path.png")
# schema_dc = nx.degree_centrality(G)
# G_undir=G.to_undirected()
# some_nodes = ['B', 'STG_C', 'C']
# G_ppt_nodes = G_undir.subgraph(some_nodes)
in_degrees_dict = dict(G.in_degree())
from random import choice
import matplotlib.pyplot as plt
import networkx as nx
from nxviz.plots import CircosPlot
G = nx.barbell_graph(m1=10, m2=3)
for n, d in G.nodes(data=True):
G.node[n]["class"] = choice(["a", "b", "c", "d", "e"])
c = CircosPlot(
G,
node_grouping="class",
group_order="alphabetically",
node_color="class",
node_order="class",
node_labels=False,
group_label_position="middle",
group_label_color=True,
)
c.draw()
d = CircosPlot(
G,
node_grouping="class",
group_order="default",
node_color="class",
node_order="class",
node_labels=False,
group_label_position="middle",
group_label_color=True,
('a', 'c', {
'class': 3
}),
('b', 'd', {
'class': 8
}),
('c', 'd', {
'class': 10
}),
('e', 'd', {
'class': 5
}),
('e', 'f', {
'class': 5
}),
]
G = nx.Graph()
G.add_nodes_from(nodelist)
G.add_edges_from(edgelist1)
c = CircosPlot(graph=G, edge_color="weight")
c.draw()
F = nx.Graph()
F.add_nodes_from(nodelist)
F.add_edges_from(edgelist2)
d = CircosPlot(graph=F, edge_color="class")
d.draw()
plt.show()
from random import choice
import matplotlib.pyplot as plt
import networkx as nx
from nxviz.plots import CircosPlot
G = nx.barbell_graph(m1=10, m2=3)
for n, d in G.nodes(data=True):
G.node[n]["class"] = choice(["a", "b", "c", "d", "e"])
c = CircosPlot(
G,
node_grouping="class",
node_color="class",
node_order="class",
node_labels=True,
group_label_position="middle",
group_label_color=True,
group_label_offset=2,
)
c.draw()
plt.show()
pos = nx.spring_layout(G) # positions for all nodes
# # First, draw nodes from table load
nx.draw_networkx_nodes(G, pos, nodelist=tl_nodes, node_color='b')
nx.draw_networkx_nodes(G, pos, nodelist=vd_nodes, node_color='r')
nx.draw_networkx_edges(G, pos, edgelist=tl_edges, edge_color='b')
nx.draw_networkx_edges(G, pos, edgelist=vd_edges, edge_color='r')
nx.draw_networkx_labels(G, pos)
nx.draw_networkx_edge_labels(G, pos)
# nx.draw_networkx_labels(G)
plt.show()
# some plots
#cpt = CircosPlot(graph=G, node_labels=True, node_color='label', edge_color='label')
#
cpt = CircosPlot(graph=G, node_labels=True)
cpt.draw()
plt.show()
nx.draw_kamada_kawai(G)
# nx.draw_networkx_labels(G)
# nx.draw_networkx_edge_labels(G)
plt.show()
# plt.savefig("path.png")
# schema_dc = nx.degree_centrality(G)
# G_undir=G.to_undirected()
# some_nodes = ['B', 'STG_C', 'C']
# G_ppt_nodes = G_undir.subgraph(some_nodes)
in_degrees_dict = dict(G.in_degree())
top50_nodes = (n for n in list(G.nodes()) if n in list(top50authors["Name"]))
G_50 = G.subgraph(top50_nodes)
for n in G_50.nodes():
G_50.node[n]["publications"] = int(
top50authors[top50authors["Name"] == n]["Count"]
)
c = CircosPlot(
G_50,
dpi=600,
node_grouping="publications",
edge_width="Count",
figsize=(20, 20),
node_color="publications",
node_labels=True,
)
c.draw()
plt.show()
## Network analysis
deg = nx.degree_centrality(G_50)
bet = nx.betweenness_centrality(G_50)
top_df = pd.DataFrame.from_dict(
[deg, bet, dict(Counter(authors_flat).most_common(50))]
).T
top_df.columns = [
"""
Shows different edge widths on CircusPlot
"""
import matplotlib.pyplot as plt
import networkx as nx
from nxviz.plots import CircosPlot
nodelist = [("a"), ("b"), ("c"), ("d"), ("e"), ("f")]
edgelist1 = [("a", "b"), ("b", "c"), ("c", "d"), ("d", "e"), ("e", "f")]
weights = list(range(6))
G = nx.Graph()
G.add_nodes_from(nodelist)
G.add_edges_from(edgelist1)
c = CircosPlot(graph=G, edge_width=weights)
c.draw()
plt.show()
"""
Displays a NetworkX lollipop graph to screen using a CircosPlot.
"""
import matplotlib.pyplot as plt
import networkx as nx
import numpy.random as npr
from nxviz.plots import CircosPlot
G = nx.lollipop_graph(m=10, n=4)
for n, d in G.nodes(data=True):
G.node[n]["value"] = npr.normal()
c = CircosPlot(G, node_color="value", node_order="value")
c.draw()
plt.show()
ws = rescale([float(G[u][v]['weight'])**70 for u, v in G.edges], 1, 50)
edgelist = [(str(u), str(v), {"weight": ws.pop(0)}) for u, v in G.edges]
# create new graph using nodelist and edgelist
g = nx.Graph(name='Protein Interaction Graph')
g.add_nodes_from(nodelist)
g.add_edges_from(edgelist)
# go through nodes in graph G and store their degree as "class" in graph g
for v in G:
g.nodes[v]["class"] = G.degree(v)
#draw
c = CircosPlot(graph=g,
figsize=(20, 20),
node_grouping="class",
node_color="class",
edge_width="weight",
node_labels=True,
fontsize=10,
fontfamily='sans-serif') #,
## group_label_offset=0.75,group_label_position='beginning',group_legend=True,
## group_label_color=True)
c.figure.tight_layout()
c.draw()
#Get the label and color for each group used by nxviz
seen = set()
colors_group = [x for x in c.node_colors
if not (x in seen or seen.add(x))] #Gets colors in RGBA
labels_group = sorted(list(set([g.nodes[n][c.node_color] for n in G.nodes])))
#Create patchList to use as handle for plt.legend()
"""
Displays a NetworkX barbell graph to screen using a CircosPlot.
"""
from nxviz.plots import CircosPlot
from random import choice
import networkx as nx
import matplotlib.pyplot as plt
G = nx.barbell_graph(m1=10, m2=3)
for n, d in G.nodes(data=True):
G.node[n]['class'] = choice(['one', 'two', 'three'])
c = CircosPlot(G, node_color="class", node_order='class')
c.draw()
plt.show()
