Skip to content

integrated_algorithms

IntegratedAlgorithm Objects

class IntegratedAlgorithm(ABC)

Abstract class modeling Memgraph's built-in graph algorithms.

These algorithms are integrated into Memgraph's codebase and are called within a relationship part of a query. For instance: MATCH p = (:City {name: "Paris"}) -[:Road * bfs (r, n | r.length <= 200 AND n.name != "Metz")]-> (:City {name: "Berlin"})

__str__

@abstractmethod
def __str__() -> str

Instance of IntegratedAlgorithm subclass is used as a string

to_cypher_lambda

@staticmethod
def to_cypher_lambda(expression: str) -> str

Method for creating a general lambda expression.

Variables r and n stand for relationship and node. The expression is used e.g. for a filter lambda, to use only relationships of length less than 200: expression="r.length < 200" with the filter lambda being: (r, n | r.length < 200)

Arguments:

  • expression - Lambda conditions or statements.

BreadthFirstSearch Objects

class BreadthFirstSearch(IntegratedAlgorithm)

Build a BFS call for a Cypher query.

The Breadth-first search can be called in Memgraph with Cypher queries such as: MATCH (a {id: 723})-[*BFS ..10 (r, n | r.x &gt; 12 AND n.y &lt; 3)]-() RETURN *; It is called inside the relationship clause, *BFS naming the algorithm, ..10 specifying depth bounds, and (r, n | &lt;expression&gt;) is a filter lambda.

__init__

def __init__(lower_bound: int = None,
             upper_bound: int = None,
             condition: str = None) -> None

Arguments:

  • lower_bound - Lower bound for path depth.
  • upper_bound - Upper bound for path depth.
  • condition - Filter through nodes and relationships that pass this condition.

__str__

def __str__() -> str

Get a Cypher query string for this algorithm.

to_cypher_bounds

def to_cypher_bounds() -> str

If bounds are specified, returns them in grammar-defined form.

DepthFirstSearch Objects

class DepthFirstSearch(IntegratedAlgorithm)

Build a DFS call for a Cypher query. The Depth-First Search can be called in Memgraph with Cypher queries such as: MATCH (a {id: 723})-[ ..10 (r, n | r.x > 12 AND n.y < 3)]-() RETURN ; It is called inside the relationship clause, "" naming the algorithm ("" without "DFS" because it is defined like such in openCypher), "..10" specifying depth bounds, and "(r, n | <expression>)" is a filter lambda.

__init__

def __init__(lower_bound: int = None,
             upper_bound: int = None,
             condition: str = None) -> None

Arguments:

  • lower_bound - Lower bound for path depth.
  • upper_bound - Upper bound for path depth.
  • condition - Filter through nodes and relationships that pass this condition.

__str__

def __str__() -> str

get Cypher query string for this algorithm.

to_cypher_bounds

def to_cypher_bounds() -> str

If bounds are specified, returns them in grammar-defined form.

WeightedShortestPath Objects

class WeightedShortestPath(IntegratedAlgorithm)

Build a Dijkstra shortest path call for a Cypher query.

The weighted shortest path algorithm can be called in Memgraph with Cypher queries such as: " MATCH (a {id: 723})-[r WSHORTEST 10 (r, n | r.weight) weight_sum (r, n | r.x > 12 AND r.y < 3)]-(b {id: 882}) RETURN * " It is called inside the relationship clause, "WSHORTEST" naming the algorithm, "10" specifying search depth bounds, and "(r, n | <expression>)" is a filter lambda, used to filter which relationships and nodes to use.

__init__

def __init__(upper_bound: int = None,
             condition: str = None,
             total_weight_var: str = DEFAULT_TOTAL_WEIGHT,
             weight_property: str = DEFAULT_WEIGHT_PROPERTY) -> None

Arguments:

  • upper_bound - Upper bound for path depth.
  • condition - Filter through nodes and relationships that pass this condition.
  • total_weight_var - Variable defined as the sum of all weights on path being returned.
  • weight_property - property being used as weight.

AllShortestPath Objects

class AllShortestPath(IntegratedAlgorithm)

Build a Dijkstra shortest path call for a Cypher query.

The weighted shortest path algorithm can be called in Memgraph with Cypher queries such as: " MATCH (a {id: 723})-[r ALLSHORTEST 10 (r, n | r.weight) total_weight (r, n | r.x > 12 AND r.y < 3)]-(b {id: 882}) RETURN * " It is called inside the relationship clause, "ALLSHORTEST" naming the algorithm, "10" specifying search depth bounds, and "(r, n | <expression>)" is a filter lambda, used to filter which relationships and nodes to use.

__init__

def __init__(upper_bound: int = None,
             condition: str = None,
             total_weight_var: str = DEFAULT_TOTAL_WEIGHT,
             weight_property: str = DEFAULT_WEIGHT_PROPERTY) -> None

Arguments:

  • upper_bound - Upper bound for path depth.
  • condition - Filter through nodes and relationships that pass this condition.
  • total_weight_var - Variable defined as the sum of all weights on path being returned.
  • weight_property - Property being used as weight.