In this lab, we shall work with Florentine families graph.
Click here to see details on Florantine family dataset. This dataset comes bundled with networkx as shown here. Visit these links to get some idea about the contents of this dataset. We shall use it as an undirected dataset for this lab and work on identifying paths between different nodes.
You will be able to:
- Load and study Florentine families graph generator in networkx
- Calculate and visualize the different paths between a given set of nodes
Note: Refer to the documentation for methods to load the required graph and layout.
# Your code here 
# Your code here ['Medici', 'Barbadori', 'Castellani', 'Peruzzi']
You will see above that the shortest paths may not necessarily be unique i.e. there can be more than one , while counting the number of hops. We can use all_shortest_paths() method to find if there are more than one.
# Your code here [['Medici', 'Barbadori', 'Castellani', 'Peruzzi'],
['Medici', 'Ridolfi', 'Strozzi', 'Peruzzi']]
- Create a function plot_paths(G, paths) that would take in a graph with calculated path(s)
- Use
fruchterman_reingold_layout() - For all paths in
paths, color the edges and display the graph - Print the path in the output
source = 'Lamberteschi'
target = 'Ridolfi'def plot_paths(G, paths):
# Your code here
pass
print(nx.shortest_path(G, source, target))
plot_paths(G, [nx.shortest_path(G, source, target)])['Lamberteschi', 'Guadagni', 'Tornabuoni', 'Ridolfi']
- Use
nx.all_simple_paths(G, source, target)to calculate all possible paths between two families and plot them.
# Your code here 1 ['Lamberteschi', 'Guadagni', 'Tornabuoni', 'Medici', 'Barbadori', 'Castellani', 'Peruzzi', 'Strozzi', 'Ridolfi']
2 ['Lamberteschi', 'Guadagni', 'Tornabuoni', 'Medici', 'Barbadori', 'Castellani', 'Peruzzi', 'Bischeri', 'Strozzi', 'Ridolfi']
3 ['Lamberteschi', 'Guadagni', 'Tornabuoni', 'Medici', 'Barbadori', 'Castellani', 'Strozzi', 'Ridolfi']
4 ['Lamberteschi', 'Guadagni', 'Tornabuoni', 'Medici', 'Ridolfi']
5 ['Lamberteschi', 'Guadagni', 'Tornabuoni', 'Ridolfi']
6 ['Lamberteschi', 'Guadagni', 'Albizzi', 'Medici', 'Barbadori', 'Castellani', 'Peruzzi', 'Strozzi', 'Ridolfi']
7 ['Lamberteschi', 'Guadagni', 'Albizzi', 'Medici', 'Barbadori', 'Castellani', 'Peruzzi', 'Bischeri', 'Strozzi', 'Ridolfi']
8 ['Lamberteschi', 'Guadagni', 'Albizzi', 'Medici', 'Barbadori', 'Castellani', 'Strozzi', 'Ridolfi']
9 ['Lamberteschi', 'Guadagni', 'Albizzi', 'Medici', 'Ridolfi']
10 ['Lamberteschi', 'Guadagni', 'Albizzi', 'Medici', 'Tornabuoni', 'Ridolfi']
11 ['Lamberteschi', 'Guadagni', 'Bischeri', 'Peruzzi', 'Castellani', 'Strozzi', 'Ridolfi']
12 ['Lamberteschi', 'Guadagni', 'Bischeri', 'Peruzzi', 'Castellani', 'Barbadori', 'Medici', 'Ridolfi']
13 ['Lamberteschi', 'Guadagni', 'Bischeri', 'Peruzzi', 'Castellani', 'Barbadori', 'Medici', 'Tornabuoni', 'Ridolfi']
14 ['Lamberteschi', 'Guadagni', 'Bischeri', 'Peruzzi', 'Strozzi', 'Castellani', 'Barbadori', 'Medici', 'Ridolfi']
15 ['Lamberteschi', 'Guadagni', 'Bischeri', 'Peruzzi', 'Strozzi', 'Castellani', 'Barbadori', 'Medici', 'Tornabuoni', 'Ridolfi']
16 ['Lamberteschi', 'Guadagni', 'Bischeri', 'Peruzzi', 'Strozzi', 'Ridolfi']
17 ['Lamberteschi', 'Guadagni', 'Bischeri', 'Strozzi', 'Castellani', 'Barbadori', 'Medici', 'Ridolfi']
18 ['Lamberteschi', 'Guadagni', 'Bischeri', 'Strozzi', 'Castellani', 'Barbadori', 'Medici', 'Tornabuoni', 'Ridolfi']
19 ['Lamberteschi', 'Guadagni', 'Bischeri', 'Strozzi', 'Peruzzi', 'Castellani', 'Barbadori', 'Medici', 'Ridolfi']
20 ['Lamberteschi', 'Guadagni', 'Bischeri', 'Strozzi', 'Peruzzi', 'Castellani', 'Barbadori', 'Medici', 'Tornabuoni', 'Ridolfi']
21 ['Lamberteschi', 'Guadagni', 'Bischeri', 'Strozzi', 'Ridolfi']
-
Modify the
plot_paths()to show each path in a different graph, with edge width showing the distance between source and target. -
Prettify the graph entities further in order to make it more presentable
In this lab, we saw how to calculate and visualize paths between a given set of nodes in a graph. The skills learned in these simple exercises can be scaled to deal with much larger networks with possible tens of thousands of nodes (or maybe more) to identify paths between node entities. We mainly looked at calculating the simple distance, but the idea can be applied to directed and weighted networks with same level of ease.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent