SET

The SET clause is used to update properties on nodes and relationships.

Set Node Properties#

Set Single Property#

MATCH (u:User {name: 'Alice'})
SET u.age = 29
RETURN u

Set Multiple Properties#

MATCH (u:User {name: 'Alice'})
SET u.age = 29, u.city = 'New York', u.updated = datetime()
RETURN u

Set from Map#

MATCH (u:User {name: 'Alice'})
SET u = {name: 'Alice', age: 29, city: 'New York'}
RETURN u

Warning: This replaces ALL properties!

Set with += (Add Properties)#

MATCH (u:User {name: 'Alice'})
SET u += {age: 29, city: 'New York'}
RETURN u

This preserves existing properties and adds/updates specified ones.

Set Relationship Properties#

MATCH (a:User)-[r:KNOWS]->(b:User)
WHERE a.name = 'Alice' AND b.name = 'Bob'
SET r.strength = 0.9, r.updated = datetime()
RETURN r

Add Labels#

Add Single Label#

MATCH (u:User {name: 'Alice'})
SET u:Premium
RETURN u

Add Multiple Labels#

MATCH (u:User {name: 'Alice'})
SET u:Premium:Verified
RETURN u

Set with Parameters#

MATCH (u:User {name: $name})
SET u.age = $age, u.city = $city
RETURN u

Python example:

graph.query("""
    MATCH (u:User {name: $name})
    SET u.age = $age
    RETURN u
""", {
    'name': 'Alice',
    'age': 29
})

Conditional Set#

Set Based on Condition#

MATCH (u:User)
SET u.status = CASE
  WHEN u.age < 18 THEN 'minor'
  WHEN u.age >= 65 THEN 'senior'
  ELSE 'adult'
END

Set If Property Exists#

MATCH (u:User)
WHERE u.tempEmail IS NOT NULL
SET u.email = u.tempEmail
REMOVE u.tempEmail

Increment/Decrement#

Increment Counter#

MATCH (p:Post {id: $postId})
SET p.views = p.views + 1
RETURN p.views

Decrement Value#

MATCH (u:User {name: 'Alice'})
SET u.credits = u.credits - 10
RETURN u.credits

Set from Expressions#

Computed Values#

MATCH (u:User)
SET u.fullName = u.firstName + ' ' + u.lastName
RETURN u

String Operations#

MATCH (u:User)
SET u.email = toLower(u.email),
    u.username = trim(u.username)
RETURN u

Date/Time Operations#

MATCH (u:User)
SET u.lastLogin = datetime(),
    u.sessionExpiry = datetime() + duration('PT1H')
RETURN u

Set All Properties#

Copy all properties from one node to another:

MATCH (source:User {name: 'Alice'}),
      (target:User {name: 'Bob'})
SET target = properties(source)
RETURN target

Set NULL (Remove Property)#

MATCH (u:User {name: 'Alice'})
SET u.tempData = NULL
RETURN u

Or use REMOVE:

MATCH (u:User {name: 'Alice'})
REMOVE u.tempData

Bulk Updates#

Update All Matching Nodes#

MATCH (u:User)
WHERE u.active = true
SET u.lastNotified = datetime()
RETURN count(u) AS updated

Update with UNWIND#

UNWIND [
  {name: 'Alice', age: 29},
  {name: 'Bob', age: 33}
] AS update
MATCH (u:User {name: update.name})
SET u.age = update.age

Common Patterns#

Update or Create Timestamp#

MATCH (u:User {name: 'Alice'})
SET u.updated = datetime()
SET u.created = coalesce(u.created, datetime())
RETURN u

Increment View Counter#

MATCH (p:Post {id: $postId})
SET p.views = coalesce(p.views, 0) + 1,
    p.lastViewed = datetime()
RETURN p

Toggle Boolean#

MATCH (u:User {name: 'Alice'})
SET u.active = NOT u.active
RETURN u.active

Add to Array#

MATCH (u:User {name: 'Alice'})
SET u.tags = coalesce(u.tags, []) + ['premium']
RETURN u

Update Based on Relationship#

MATCH (u:User)-[:POSTED]->(p:Post)
WITH u, count(p) AS postCount
SET u.totalPosts = postCount
RETURN u

Normalize Data#

MATCH (u:User)
SET u.email = toLower(trim(u.email)),
    u.name = trim(u.name)
WHERE u.email IS NOT NULL
RETURN count(u) AS normalized

Set Default Values#

MATCH (u:User)
WHERE u.role IS NULL
SET u.role = 'user'
RETURN count(u) AS updated

SET with CREATE#

CREATE (u:User)
SET u.name = 'Alice',
    u.created = datetime(),
    u.id = randomUUID()
RETURN u

SET with MERGE#

MERGE (u:User {email: $email})
ON CREATE SET 
  u.name = $name,
  u.created = datetime()
ON MATCH SET
  u.lastSeen = datetime()
RETURN u

Performance Tips#

Batch Updates ```cypher // Good: Single query updates many MATCH (u:User) WHERE u.active = true SET u.updated = datetime()

// Bad: Many individual queries // Multiple SET queries for each user ```

Use += for Partial Updates ```cypher // Good: Only updates specified properties SET u += {age: 29}

// Bad: Replaces all properties SET u = {age: 29} ```

Index Updated Properties ```cypher CREATE INDEX ON :User(lastLogin)

// Fast queries on updated property MATCH (u:User) WHERE u.lastLogin > datetime() - duration('P7D') RETURN u ```

Minimize Property Writes ```cypher // Good: Only update if changed MATCH (u:User {name: 'Alice'}) WHERE u.age <> 29 SET u.age = 29

// Bad: Always writes even if unchanged MATCH (u:User {name: 'Alice'}) SET u.age = 29 ```

Common Mistakes#

Overwriting All Properties#

// WRONG: Removes all other properties
MATCH (u:User {name: 'Alice'})
SET u = {age: 29}  // name is now gone!

// CORRECT: Use +=
MATCH (u:User {name: 'Alice'})
SET u += {age: 29}  // name is preserved

Not Handling NULL#

// WRONG: Error if credits is NULL
SET u.credits = u.credits + 10

// CORRECT: Handle NULL
SET u.credits = coalesce(u.credits, 0) + 10

Next Steps#

Create data with CREATE
Delete data with DELETE
Match patterns with MATCH
Read Cypher Introduction

Edit this page on GitHub