Using Pandas data frames

If your results are a bit more complex and you like working with Pandas data frames, you can use the keys returned by execute_query() to easily load the results into a data frame.

For example, to get the top 1000 ASes from AS Rank with their name as known by RIPE and their registered country according to the NRO delegated stats:

import pandas as pd

# [...] database init like above

query = """
    MATCH (a:AS)-[r:RANK {reference_name: 'caida.asrank'}]-(:Ranking)
    OPTIONAL MATCH (a)-[:NAME {reference_name: 'ripe.as_names'}]-(n:Name)
    OPTIONAL MATCH (a)-[:COUNTRY {reference_name: 'nro.delegated_stats'}]-(c:Country)
    RETURN a.asn AS asn, r.rank AS as_rank, n.name AS name, c.country_code as country
    ORDER BY as_rank
    LIMIT 1000
"""

records, _, keys = db.execute_query(query)
df = pd.DataFrame(records, columns=keys)

# Alternatively, Neo4j can directly transform the results into a dataframe
# df = db.execute_query(query, result_transformer_=neo4j.Result.to_df)

For querying IYP this is pretty much all you need to know. For more examples see the Jupyter notebooks we provide as part of our paper. However, you will notice that only the queries are more involved, the Python functions are the same.

References

Simple Python example Hosting a local IYP instance