How to save a 'pos dictionary' type of object? (NetworkX)

Question

I am using the networkx as nx. After creating a graph object (G), I use the data = nx.spring_layout(G) method to generate a dictionary (pos). Generating this pos object is very time-consuming and I want to save it as a file, so I don't have to rerun it all whenever I want to work with this dataset. This is what it looks like so far:

G = nx.Graph()
G.add_edges_from(tuples)
data = nx.spring_layout(G)

If I run type(data), I receive 'dict'. If I run type(G), I receive 'networkx.classes.graph.Graph'.

Since it is a 'dict' object, these are some solutions I found online that I tried and didn't work:

1. Save it as a .csv file

import csv

w = csv.writer(open("output.csv", "w"))

for key, val in data.items():
    w.writerow([key, val])

2. Save as JSON file

import json
json = json.dumps(data)


f = open("data.json","w")

f.write(json)

f.close()

3. Save as a .txt file

f = open("data.txt","w")

f.write( str(data) )

f.close()

Hopefully I made it clear that (i) I don't know which type of file is the best one to save this sort of object; that (ii) I don't know how to save it in any way; and (iii) that, naturally, I don't know how to properly load this file as 'pos dict' object once it is saved in my directory.

For reference, nx.spring_layout(G) returns: "pos: dict A dictionary of positions keyed by node.", as described in the documentation.

Answer 1

Probably the simplest way to store a python object for later use is to use the pickle module:

import pickle

# save data
with open('positions.pickle', 'wb') as f:
    pickle.dump(data, f)

# load data
with open('positions.pickle', 'rb') as f:
    data = pickle.load(f)