client.dag

The DAG captures the interconnected graph of tasks and their dependencies.

Attributes

logger

Classes

Dag

The DAG captures the interconnected graph of tasks and their dependencies.

Module Contents

client.dag.logger[source]

class client.dag.Dag(requester: jobmon.core.requester.Requester | None = None)[source]

Bases: object

The DAG captures the interconnected graph of tasks and their dependencies.

nodes: Set[jobmon.client.node.Node][source]

requester = None[source]

property dag_id: int[source]: Database unique ID of this DAG.

add_node(node: jobmon.client.node.Node) → None[source]

Add a node to this dag.

Parameters:: node (jobmon.client.node.Node) – Node to add to the dag

bind(chunk_size: int = 500) → int[source]

Retrieve an id for a matching dag from the server.

If it doesn’t exist, first create one, including its edges.

validate() → None[source]: Validate the nodes and their dependencies.

_is_cyclic(dag_map: Dict[jobmon.client.node.Node, Set[jobmon.client.node.Node]]) → bool[source]

Return true if the nodes are cyclic.

This method is effectively a depth-first search looking for already-seen nodes, implemented using the “stack of iterators” pattern to get around Python’s recursion limit of 1000.

_bulk_bind_nodes(chunk_size: int) → None[source]

_get_dag_id() → int | None[source]

_bulk_insert_edges(dag_id: int, chunk_size: int = 500) → None[source]

__hash__() → int[source]: Determined by hashing all sorted node hashes and their downstream.

__repr__() → str[source]: A representation string for a Dag instance.