#!/usr/bin/env python
# coding: utf-8

# # Panama Papers Explorer - py2neo
# 
# This notebook provides a quick start for using py2neo with the Panama Papers neo4j docker container [ryguyrg/neo4j-panama-papers](https://hub.docker.com/r/ryguyrg/neo4j-panama-papers/) [[Github](https://github.com/ryguyrg/panama-neo4j)].
# 
# The following `docker-compose.yml` file will lauch a Jupyter notebook environment linked to the neo4j container the Panama Papers data.

# ````
# #Run with:
# #docker-compose up
# #Then visit IP_ADDRESS:7474
# #The defaul credentials are neo4j/neo4j
# #You will be prompted for a new password - I used: panamapapers
# 
# #Get the Panama Papers neo4j container
# neo4j:
#   image: ryguyrg/neo4j-panama-papers
#   ports:
#     - "7474:7474"
#     - "1337:1337"
#   volumes:
#     - /opt/data
# 
# #Download a Jupyter notebook environment and link the neo4j container to it
# jupyterscipy:
#   image: jupyter/scipy-notebook
#   ports:
#     - "8890:8888"
#   links:
#     - neo4j:neo4j
#   volumes:
#     - ./notebooks:/home/jovyan/work
# 
# ##If you prefer an RStudio environment:
# #rstudio:
# #  image: rocker/rstudio 
# #  ports:
# #    - "8787:8787"
# #  links:
# #    - neo4j:neo4j
# #  volumes:
# #    - ./rstudio:/home/rstudio
# ##Then install: install.packages('RNeo4j', repos="http://cran.rstudio.com/")
# ````

# ## py2neo
# 
# `py2neo` provides a Python wrapper for neo4j.

# In[1]:


get_ipython().system('pip3 install py2neo')
from py2neo import Graph


# Create a connection to the `neo4j` database, using the `neo4j` user account and your new password:

# In[2]:


graph = Graph("http://neo4j:7474/db/data/",user='neo4j',password='panamapapers')


# Start to explore the graph - what node types are there?

# In[3]:


graph.node_labels


# For each node type, what attributes/labels are there?

# In[4]:


for nl in graph.node_labels:
    print(nl)
    for i in graph.schema.get_indexes(nl):
        print('\t{}'.format(i))


# What relationship types are there?

# In[5]:


graph.relationship_types


# ## Officer Searches
# 
# Some simple searches on Officers.

# In[62]:


#Find officers whose name partially string matches the supplied name
name='Smith'
for r in graph.run("MATCH (a:Officer) WHERE a.name CONTAINS '{}' RETURN a.name LIMIT 10".format(name)):
    print(r)


# In[99]:


#Return the fill record for officers whose name partially string matches the supplied name
#Make case insenstive by casting all parts to uppercase
exact_name='David Smith'
for r in graph.run("MATCH (a:Officer) WHERE UPPER(a.name) ='{}' RETURN a LIMIT 10".format(name.upper())):
    print(r)


# In[63]:


p=graph.find_one('Officer','name',exact_name)
p


# In[80]:


#Find what things a person is connected to at the node level as a parent
for c in graph.match(start_node=p):
    print(c)
    print(c.start_node())
    print(c.type())
    print(c.end_node())
    print('----')


# In[43]:


#What companies is the person associated with in more detail for a specified relationship type?
q='''
MATCH (o:Officer), (e:Entity)
WHERE o.name='{}' AND (o)-[:SHAREHOLDER_OF]-(e)

RETURN o,e LIMIT 10
'''.format(name)

for r in graph.run(q):
    print(r)


# In[46]:


exact_entity_name='PORTONES DEL MAR INVESTORS  INC.'
e=graph.find_one('Entity','name',exact_entity_name)
e


# In[54]:


#Find what things a company is connected to at the node level - company is an "end_node"
for c in graph.match(end_node=e,limit=10):
    print(c)


# In[105]:


iid=e['internal_id']
for r in graph.run("MATCH (a:Entity) WHERE a.internal_id ='{}' RETURN a LIMIT 10".format(iid)):
    print(r)


# In[ ]: