Autolayout graph nodes for interrelation diagram

diagram.svg

interrelationdiagram.py

from igraph import *
import numpy as np

# Data
edges = [(0,6),(0,7),(0,16),(0,19),(1,0),(1,3),(1,5),(1,16),(2,1),(2,3),(2,5),(2,11),(2,13),(2,14),(2,15),(2,16),(2,20),(2,22),(3,4),(3,12),(3,14),(3,16),(3,22),(4,22),(5,1),(5,2),(5,3),(5,6),(5,10),(5,11),(5,13),(5,14),(5,16),(5,20),(5,22),(5,23),(6,7),(6,19),(7,0),(7,4),(7,16),(7,18),(7,19),(8,0),(8,9),(8,16),(8,22),(9,0),(9,16),(10,0),(10,6),(11,0),(11,1),(11,3),(11,4),(11,7),(11,10),(11,13),(11,14),(11,16),(11,22),(12,0),(12,1),(12,6),(12,10),(12,11),(12,13),(12,14),(12,16),(12,20),(12,21),(12,22),(13,0),(13,6),(13,9),(13,14),(13,16),(13,18),(14,0),(14,1),(14,6),(14,9),(14,10),(14,20,),(14,22),(15,0),(15,3),(15,5),(15,8),(15,9),(15,12),(15,13),(15,16),(15,20),(15,21),(15,22),(15,23),(16,6),(16,7),(16,10),(16,14),(16,20),(16,22),(17,0),(17,2),(17,3),(17,8),(17,9),(17,10),(17,15),(17,22),(18,0),(18,11),(18,16),(18,19),(19,11),(20,6),(20,7),(20,19),(21,8),(21,9),(21,22),(22,4),(22,9),(22,19),(23,20)]

labels = ["a", "b", "c", "d", "e", "f", "g", "h", "i", "j", "k", "l", "m", "n", "o", "p", "q", "r", "s", "t", "u", "v", "w", "x"]

N = len(labels)

# Define colors used for outdegree visualization
colours = ['#ffffb2', '#fecc5c', '#fd8d3c', '#e31a1c']

# Graph generation

g = Graph(vertex_attrs={"label": labels}, edges=edges, directed=True)

# Add in/outdegree to label
g.vs["label"] = [l+"\n\nIn: "+str(g.indegree(i))+"    Out: "+str(g.outdegree(i)) for i,l in enumerate(g.vs["label"])]

# Calculate outdegree
degrees = g.outdegree()

# Order vertices in bins based on outdegree
bins = np.linspace(0, max(degrees), len(colours))  
digitized_degrees =  np.digitize(degrees, bins)

# Set colors according to bins
g.vs["color"] = [colours[x-1] for x in digitized_degrees]

# Also color the edges
for ind, color in enumerate(g.vs["color"]):
        edges = g.es.select(_source=ind)
        edges["color"] = [color]

# Community detection
communities = g.community_edge_betweenness(directed=True)
clusters = communities.as_clustering()

# Alternative community detection
#clusters = g.community_infomap(vertex_weights=degrees, trials=1)

# Set edge weights based on communities
weights = {v: len(c) for c in clusters for v in c}
g.es["weight"] = [weights[e.tuple[0]] + weights[e.tuple[1]] for e in g.es]

# Set visual style for the plot
visual_style = {}
# Choose the layout
visual_style["layout"] = g.layout_fruchterman_reingold(weights=g.es["weight"], maxiter=1000, area=N**3, repulserad=N**3)

# Other options
visual_style["vertex_shape"] = "rectangle"
visual_style["vertex_size"] = 85
visual_style["vertex_label_dist"] = 0
visual_style["vertex_label_size"] = 8
visual_style["vertex_frame_color"] = g.vs["color"]
visual_style["vertex_label_family"] = "Calibri"
visual_style["wrap_labels"] = True
visual_style["bbox"] = (764,875)
visual_style["margin"] = 50

# Plot the graph
plot(g, target="diagram.svg", **visual_style)