compute.delayed

cloud.compute.delayed

Classes

Name	Description
Delayed	Serverless Python function execution as a future.
DelayedArrayUDF	Serverless Array UDF called with the
DelayedBase	Base Delayed interface class
DelayedMultiArrayUDF	Serverless Multi-Array UDF called with the
DelayedSQL	Serverless SQL query on a registered Array called with the

Delayed

cloud.compute.delayed.Delayed(
    self
    func_exec
    *args
    local=False
    mode=Mode.REALTIME
    **kwargs
)

Serverless Python function execution as a future.

Extends DelayedBase

DelayedArrayUDF

cloud.compute.delayed.DelayedArrayUDF(self, uri, func_exec, *args, **kwargs)

Serverless Array UDF called with the Delayed API

Extends DelayedBase

DelayedBase

cloud.compute.delayed.DelayedBase(
    self
    func
    *args
    name=None
    dag=None
    local_mode=False
    mode=Mode.REALTIME
    **kwargs
)

Base Delayed interface class

Methods

Name	Description
all	Run a list of Delayed object all in parallel
compute	Starts execution of all Delayed tasks associated with this node.
visualize	Build and render a tree diagram of the DAG.

all

cloud.compute.delayed.DelayedBase.all(futures, namespace=None)

Run a list of Delayed object all in parallel

Parameters

Name	Type	Description	Default
futures		list of Delayed objects to run	required
namespace		optional namespace to run all tasks in	None

Returns

Name	Type	Description
		list of results in order of futures

compute

cloud.compute.delayed.DelayedBase.compute(
    namespace=None
    name=None
    max_workers=None
)

Starts execution of all Delayed tasks associated with this node.

Parameters

Name	Type	Description	Default
namespace	Optional[str]	The namespace to execute tasks under, if different than the user’s default.	None
name	Optional[str]	An optional name to identify the task graph in logs.	None

Returns

Name	Type	Description
	Any	results

visualize

cloud.compute.delayed.DelayedBase.visualize(
    notebook=True
    auto_update=True
    force_plotly=False
)

Build and render a tree diagram of the DAG.

Parameters

Name	Type	Description	Default
notebook		Is the visualization inside a jupyter notebook? If so we’ll use a widget	True
auto_update		Should the diagram be auto updated with each status change	True
force_plotly		Force the use of plotly graphs instead of TileDB Plot Widget	False

Returns

Name	Type	Description
		returns plotly figure

DelayedMultiArrayUDF

cloud.compute.delayed.DelayedMultiArrayUDF(
    self
    func
    array_list
    *args
    **kwargs
)

Serverless Multi-Array UDF called with the Delayed API

Extends DelayedBase

DelayedSQL

cloud.compute.delayed.DelayedSQL(self, *args, **kwargs)

Serverless SQL query on a registered Array called with the Delayed API

Extends DelayedBase