euanong/cypher_kernel

An experimental Jupyter kernel wrapping the Cypher query language and Neo4j via a REPLWrapper around cypher-shell.

Cypher Kernel

This is a small Jupyter kernel wrapping the Cypher language and Neo4j https://neo4j.com/developer/cypher/.

Why? Do I need a Cypher language kernel?

I started working on this kernel as I am teaching on Cypher and Neo4j as part of a database course on Copenhagen Business Academy.

Usually, my lecture notes are in Jupyter notebooks and I use the Jupter extension RISE to make slideshows with executable code.

In the classes on graph databases, I want to focus on the query language Cypher. That is, I do not want to have Cypher embedded in Python such as with icypher -a %cypher magic sending queries with the help of py2neo- or in plain Python notebooks

How does it look like?

In case you do not have a Python environment with Jupyter installed on your system, you can see in the following how the rendered notebooks -not the presentations- could look like, see:

• https://nbviewer.jupyter.org/github/HelgeCPH/cypher_kernel/blob/master/example/paradise_papers.ipynb
• https://nbviewer.jupyter.org/github/HelgeCPH/cypher_kernel/blob/master/example/Mark_Needhams_Examples.ipynb
• https://nbviewer.jupyter.org/github/HelgeCPH/cypher_kernel/blob/master/example/movie_graph.ipynb or
• http://htmlpreview.github.io/?https://github.com/HelgeCPH/cypher_kernel/blob/master/example/movie_graph.html

Installation

To install the cypher_kernel from PyPi:

pip install cypher_kernel
python -m cypher_kernel.install

To work on this code directly, you may want to:

git clone [email protected]:HelgeCPH/cypher_kernel.git
cd cypher_kernel
pip install .
python -m cypher_kernel.install

Configuration

To configure a Neo4j user, password, and the address of the Neo4j REST API, you can specify the values in the configuration file cypher_config.yml. Normally, this file is located under ~/.jupyter/. In case the file is not existent the following default configuration is used:

user: 'neo4j'
pwd: 'neo4j'
host: 'localhost:7474'
connect_result_nodes: False
cmd_timeout: null

Using the Cypher Kernel

Notebook: The New menu in the notebook should show an option for an Cypher notebook.

Console frontends: To use it with the console frontends, add --kernel cypher to their command line arguments.

Neo4j for Presentations

To get quickly started -under the assumption you have Docker installed- start up a Neo4j DBMS instance with:

docker run --rm --publish=7474:7474 --publish=7687:7687 neo4j
docker run --rm --publish=7474:7474 --publish=7687:7687 --env=NEO4J_dbms_memory_pagecache_size=4G neo4j

• Navigate with your browser to http://localhost:7474
• Login with neo4j as username and password respectively
• Change the password to a new one. OBS Do not forget to add this password to the cypher_config.yml, see above.
• Now, create a new Cypher notebook.

See more on configuring the Neo4j container https://neo4j.com/docs/operations-manual/current/installation/docker/

What? I have Docker but no pip and other Python stuff?!

Likely the easiest way to get started, have:

• A Linux/OS X (Windows should work but I cannot test it at the moment...)
• An installation of Anaconda (with Python 3.6), see https://www.anaconda.com/download/. Download and install it according to their documentation, see https://docs.anaconda.com/anaconda/install/
• A Docker installation, see https://www.docker.com/community-edition#/download. Alternatively, a native installation of Neo4j, see https://neo4j.com/download/
• The cypher_kernel:

pip install cypher_kernel
python -m cypher_kernel.install

• The Jupyter Notebook server up and running:

jupyter notebook

That should be it...