Retesting after refactoring
After you have refactored the graph, you should revisit all queries for your use cases.
You should first determine if any of the queries need to be rewritten to take advantage of the refactoring.
Next, we rewrite some of our queries to take advantage of the refactoring.
During your testing on your real application and especially with a fully-scaled graph, you can also profile the new queries to see if it improves performance. On the small instance model we are using, you will not see significant improvements, but you may see differences in the number of rows retrieved.
Use case #1: What people acted in a movie?
We rewrite this query to use the Actor label.
Original code:
MATCH (p:Person)-[:ACTED_IN]-(m:Movie)
WHERE m.title = 'Sleepless in Seattle'
RETURN p.name AS ActorNew code:
MATCH (p:Actor)-[:ACTED_IN]-(m:Movie)
WHERE m.title = 'Sleepless in Seattle'
RETURN p.name AS ActorProfiling Queries
For the query that uses the Person label we see this result that first retrieves the 5 Person nodes:
For the query that uses the Actor label we see this result that first retrieves the 4 Actor nodes, a slight improvement for this small graph:
Use case #3: What movies did a person act in?
We rewrite this query to use the Actor label.
Original code:
MATCH (p:Person)-[:ACTED_IN]-(m:Movie)
WHERE p.name = 'Tom Hanks'
RETURN m.title AS MovieNew code:
MATCH (p:Actor)-[:ACTED_IN]-(m:Movie)
WHERE p.name = 'Tom Hanks'
RETURN m.title AS MovieUse case #5: Who was the youngest person to act in a movie?
We rewrite this query to use the Actor label.
Original code:
MATCH (p:Person)-[:ACTED_IN]-(m:Movie)
WHERE m.title = 'Hoffa'
RETURN p.name AS Actor, p.born as `Year Born` ORDER BY p.born DESC LIMIT 1New code:
MATCH (p:Actor)-[:ACTED_IN]-(m:Movie)
WHERE m.title = 'Hoffa'
RETURN p.name AS Actor, p.born as `Year Born` ORDER BY p.born DESC LIMIT 1Use case #6: What role did a person play in a movie?
We rewrite this query to use the Actor label.
Original code:
MATCH (p:Person)-[r:ACTED_IN]-(m:Movie)
WHERE m.title = 'Sleepless in Seattle' AND
p.name = 'Meg Ryan'
RETURN r.role AS RoleNew code:
MATCH (p:Actor)-[r:ACTED_IN]-(m:Movie)
WHERE m.title = 'Sleepless in Seattle' AND
p.name = 'Meg Ryan'
RETURN r.role AS RoleUse case #8: What drama movies did an actor act in?
We rewrite this query to use the Actor label.
Original code:
MATCH (p:Person)-[:ACTED_IN]-(m:Movie)
WHERE p.name = 'Tom Hanks' AND
'Drama' IN m.genres
RETURN m.title AS MovieNew code:
MATCH (p:Actor)-[:ACTED_IN]-(m:Movie)
WHERE p.name = 'Tom Hanks' AND
'Drama' IN m.genres
RETURN m.title AS MovieUse case #10: What actors were born before 1950?
For completeness of our testing, we add this new use case.
We rewrite this query to use the Actor label.
Original code:
MATCH (p:Person)
WHERE p.born < '1950'
RETURN p.nameNew code:
MATCH (p:Actor)
WHERE p.born < '1950'
RETURN p.nameProfile queries
If you have a scaled graph (with more nodes/relationships than what we have been using in this course), you should also use the PROFILE keyword to compare the performance of the queries after the refactoring.
Check your understanding
Steps after refactoring
After you have refactored the graph, what must you do?
-
❏ Test the consistency of the graph.
-
✓ Rewrite any Cypher queries for use cases that are affected by the refactoring.
-
❏ CALL
db.cleanup()to make sure all orphaned pointers are deleted. -
✓ Retest all use cases that are affected by the refactoring.
Hint
You must do these two things after every refactoring of the graph.
Solution
After refactoring the graph, you must rewrite any Cypher queries and retest all use cases that have been affected as an outcome of the refactoring.
Summary
In this lesson, you learned that you may need to rewrite your Cypher that supports your use cases before you retest after refactoring. In the next challenge, you will retest all use cases because you have added the Actor label to the graph.