README

Summary

This is an example project combining Neo4j, Python and Github API

This project demostrates:

1. Call Github API using Python multithreading

2. Collect data from threads and output to a followers.csv

3. Use followers.csv which contains User data and Follower relationships to output CSVs distict for each:

   • users.csv - distinct Users
   • follows.csv - distinct mappings from User-to-Follows

4. Load CSVs into Neo4j

5. Create Neo4j relationships

6. Query Neo4j using the Python neo4j-driver client

7. Write queried data to write json

8. Use json to generate a D3Js web graph

Quick Example

The current web/d3-results.json is already populated with my data, so just run:

cd web
npm install http-server
http-server

To build and run the project with your data

Create a Github API token - this will be needed to make requests if making over 60 requests an hour, the unauthenticated rate limit

pip install -r requirements.txt

Populate CSVs for the Github User whose followers is being populated

python code/followers.py --username <username>
python code/follows.py --username <username>
python code/users.py --username <username>

Data can be loaded into Neo4j using the below queries

After data is loaded, run this command to populate web/d3-results.json.

Environment varialbe NEO4J_PASSWORD will need to be set for the Neo4j database connection in query.py

python query.py

Start Node server and see data with D3Js

cd web
npm install http-server
http-server

Neo4j Queries

Load CSV

CSV must first be placed in the Neo4j /import/ dir:

/Users/aaron/Library/Application\ Support/Neo4j\ Desktop/Application/neo4jDatabases/database-0bf7614b-cb8d-405e-baef-e1b94485e662/installation-3.3.5/import/

Load Users table using CSV:

LOAD CSV WITH HEADERS FROM "file:///users.csv" AS row
CREATE (u:User)
SET u = row

Load Follows table using CSV:

LOAD CSV WITH HEADERS FROM "file:///follows.csv" AS row
CREATE (u:Follows)
SET u = row

Create Relationships

Create FOLLOWS relationships using previous 2 tables:

MATCH (u:User),(f:Follows), (u2:User)
WHERE u.login = f.login
  AND f.follows = u2.login
CREATE (u)-[:FOLLOWS]->(u2)

If you have to delete data and start over

First delete relationships:

MATCH ()-[r:FOLLOWS]-() 
DELETE r

Then delete tables:

MATCH (r:User)
DELETE r

MATCH (r:Follows)
DELETE r

Libraries Used

Database

Neo4j

Python

requests

neo4j-driver

Javascript

d3js

queue.js

LICENSE

BSD-2

CONTRIBUTING

Contact me if you found this interesting

FUTURE PLANS

Add more end-to-end Neo4j Graph Database examples

Example Graph output images from this repo

1st and 2nd level relationships (D3Js)

1st level relationships (Neo4j)

1st and 2nd level relationships (Neo4j)