liveBook · Manning

Graph pattern

As explained in Graph Algorithms for Data Science

A graph pattern is a Cypher syntax representation of nodes and relationships that describes a portion of graph structure.

A graph pattern is Cypher syntax that describes nodes, relationships, and the structure connecting them. Node patterns use parentheses; labels, properties, and variables can narrow the nodes of interest. Relationship patterns connect node patterns and can include a type and direction. The same description can represent data to find, create, merge, or delete. section 3.1.1 section 3.1.2 section 3.1.6

For example, `(p:Person {name:'Satish'})` describes a Person node whose name is Satish. In `MATCH`, this is a request to retrieve matching stored data rather than instructions for how to retrieve it: Cypher is declarative. Labels and key properties can appear inline in the pattern, while additional filters can be placed in `WHERE`. Both approaches can produce the same query plan and results. section 3.1.2

Listing 3.6 Cypher statement that searches and retrieves any nodes with a label Person and a name property with a value of Satish
MATCH (p:Person {name:'Satish'})
RETURN p

This query uses an inline node pattern to retrieve the Person node named Satish, then returns the node through its variable `p`. section 3.1.2

Choose the clause according to what the pattern should do:

  • Use `CREATE` to store the described nodes and relationships. It creates the pattern each time and does not check whether matching data already exists. section 3.1.1
  • Use `MATCH` to retrieve an existing pattern. A relationship type may be omitted when any relationship type between the described nodes is acceptable. section 3.1.2 section 3.1.6
  • Use `MERGE` to match the full inline pattern first and create it only when absent. This supports repeatable imports, but matching on extra properties can produce a distinct node pattern. section 3.1.7
  • Use `MATCH` followed by `DELETE` to remove a matched pattern. `DETACH DELETE` removes a node together with attached relationships. section 3.1.6

Count hashtag pairs in the same tweet

A longer pattern can express a shared context. `MATCH (h1:Tag)<-[:HAS_TAG]-()-[:HAS_TAG]->(h2:Tag)` finds two Tag nodes attached to one tweet. Counting those matches measures co-occurrence. Because each pair can appear in both variable orders, `WHERE id(h1) < id(h2)` removes duplicate pairings before the count is returned. section 7.2

Patterns also support exploratory analysis of a Twitter network: they can select users, tweets, and relationships before aggregation, subqueries, or cardinality handling. For co-occurrence projections, a conceptually undirected connection may be represented as two opposite directed relationships when working with Graph Data Science. section 4.0 section 7.2

How Graph Algorithms for Data Science covers it

How deep it goes
Explains inline labels and properties, WHERE filters, clause behavior, and a multi-hop co-occurrence query without formal graph theory.
Examples
Demonstrates Person nodes, FRIEND relationships, deletion targets, and hashtag pairs sharing a tweet.
What it assumes
Teaches pattern use through basic Person-node examples and assumes a Neo4j database with imported Twitter data for exploratory analysis.
What you can do after
Apply it via Cypher clauses to describe graph structures for retrieval, updates, deletion, and co-occurrence counts.
How it connects
Connects MATCH, WHERE, CREATE, MERGE, DELETE, OPTIONAL MATCH, aggregation, and Graph Data Science relationship representation.
Cover of Graph Algorithms for Data Science
liveBookmanning.com
Read Graph Algorithms for Data Science on liveBook Get the book at manning.com

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Graph Algorithms for Data Science. The section links point to the parts of the book it draws on.