apoc.export.arrow.stream.graph

Procedure

apoc.export.arrow.stream.graph(graph ANY, config MAP<STRING, ANY>) - exports the given graph as an arrow byte array.

It is considered unsafe to run this procedure using multiple threads. It is therefore not supported by the parallel runtime (introduced in Neo4j 5.13). For more information, see the Cypher Manual → Parallel runtime.

s == Signature

apoc.export.arrow.stream.graph(graph :: ANY, config = {} :: MAP) :: (value :: LIST<INTEGER>)

Input parameters

Name Type Default

graph

ANY

null

config

MAP

null

Output parameters

Name Type

value

LIST<INTEGER>

Config parameters

The procedure support the following config parameters:

Table 1. Config parameters
name type default description

batchSize

Integer

2000

the batch size of the ArrowStreamWriter

Usage Examples

The procedure expose an Arrow byte[] for each batch of rows with the following structure - <id>: for node id - <labels>: list of labels - <source.id>: source node id (in case of relationship) - <target.id>: target node id (in case of relationship) - <type>: for relationship type - the list of properties of nodes and relationships flattened as table

So for the following query:

CREATE (f:User {name:'Adam',age:42,male:true,kids:['Sam','Anna','Grace'], born:localdatetime('2015185T19:32:24'), place:point({latitude: 13.1, longitude: 33.46789})})-[:KNOWS {since: 1993, bffSince: duration('P5M1.5D')}]->(b:User {name:'Jim',age:42}),(c:User {age:12}),(d:Another {foo: 'bar'})

With this query:

CALL apoc.graph.fromDB('neo4j',{}) yield graph
CALL apoc.export.arrow.stream.graph(graph)
YIELD value RETURN value

We’ll have a table with the following columns:

  • <id>

  • <labels>

  • <source.id>

  • <target.id>

  • <type>

  • name

  • age

  • male

  • kids

  • born

  • place

  • since

  • bffSince

More documentation of apoc.export.arrow.stream.graph