Ctrl+K

Hello World Neo4j

Hello World Neo4j#

This notebook walks through basic code examples of integrating xGT with Neo4j, pandas, networkx, and jgraph

trovares_connector#

trovares_connector can be installed with: python -m pip install trovares_connector

Connect#

Connect to xGT and the Neo4j database

from trovares_connector import Neo4jConnector, Neo4jDriver
import xgt

xgt_server = xgt.Connection()
xgt_server.set_default_namespace('neo4j')
database = "test"
neo4j_server = Neo4jDriver(auth=('neo4j', 'foo'), database=database)

c=Neo4jConnector(xgt_server, neo4j_server)

neo4j_server.query("MATCH (n) DETACH DELETE n").finalize()

Create a graph in Neo4j#

names = ["Nicole", "Nathan", "Bishop", "Ashley", "Joe"]
neo4j_server.query(f'create (a:Person{{id:0, name:"{names[0]}"}})').finalize()
for i in range(0, len(names) - 1):
    neo4j_server.query(f'match(a:Person) where a.id = {i} create (a)-[:Knows]->(:Person{{id:{i + 1},name:"{names[i + 1]}"}})').finalize()
neo4j_server.query('match(a:Person), (b:Person) where a.id = 2 and b.id = 0 create (a)-[:Knows]->(b)').finalize()

Get Neo4j schema info#

c.neo4j_relationship_types

['Knows']

c.neo4j_node_labels

['Person']

c.neo4j_edges

{'Knows': {'schema': {},
  'endpoints': {('Person', 'Person')},
  'sources': {'Person'},
  'targets': {'Person'}}}

c.neo4j_nodes

{'Person': {'name': 'String', 'id': 'Long'}}

Transfer data to xGT#

Transfers data associated with edge, Knows, including the associated vertices

c.transfer_to_xgt(edges=["Knows"])

Transferring: [############################################################] 10/10 in 0:00:00.3s (31.5/s, eta: 0:00:00.0s)

Cypher#

Retrieve python query results

query = "match(a)-->()-->()-->(a) return a.name"

job = xgt_server.run_job(query)
print("xGT found the following relationship triangle: " + ', '.join(str(row[0]) for row in job.get_data()))

xGT found the following relationship triangle: Bishop, Nathan, Nicole

Save results to disk#

query = "match(a)-->()-->()-->(a) return a.name into results"

xgt_server.run_job(query)
xgt_server.get_frame("results").save("results.csv")
xgt_server.drop_frame("results")

True

Pandas DataFrames#

Convert results into a pandas dataframe

job.get_data_pandas()
a_name
0 Bishop
1 Nathan
2 Nicole

NetworkX#

Getting data into networkx

import networkx as nx

edge_data = [tuple(x) for x in xgt_server.get_frame("Knows").get_data()]
vertex_data = xgt_server.get_frame("Person").get_data()

dg = nx.DiGraph()

dg.add_edges_from(edge_data)
nx.set_node_attributes(dg, {row[0]:row[1] for row in vertex_data}, name="name")

nx.draw_networkx(dg, font_color='white')
../../_images/c73ee4e53e49e2a6adbafa2759015d0b6f18c4249decb772f54f45998d8d4c03.png 
dg.nodes(data=True)

NodeDataView({0: {'name': 'Nicole'}, 1: {'name': 'Nathan'}, 2: {'name': 'Bishop'}, 3: {'name': 'Ashley'}, 4: {'name': 'Joe'}})

nx.degree(dg)

DiDegreeView({0: 2, 1: 2, 2: 3, 3: 2, 4: 1})

jgraph#

Visualizing the graph with jgraph

import jgraph

data =  xgt_server.run_job("match(a)-->(b) return a.id, b.id").get_data()
data = [tuple(x) for x in data]
jgraph.draw(data)
Contents