nipype.pipeline.plugins.base module

Common graph operations for execution.

class nipype.pipeline.plugins.base.DistributedPluginBase(plugin_args=None)

Bases: nipype.pipeline.plugins.base.PluginBase

Execute workflow with a distribution engine

Combinations of proc_done and proc_pending:

proc_done

proc_pending

outcome

True

False

Process is finished

True

True

Process is currently being run

False

False

Process is queued

False

True

INVALID COMBINATION

procs

list (N) of underlying interface elements to be processed

Type:

list

proc_done

a boolean numpy array (N,) signifying whether a process has been submitted for execution

Type:

numpy.ndarray

proc_pending

a boolean numpy array (N,) signifying whether a process is currently running.

Type:

numpy.ndarray

depidx

a boolean matrix (NxN) storing the dependency structure across processes. Process dependencies are derived from each column.

Type:

numpy.matrix

run(graph, config, updatehash=False)

Executes a pre-defined pipeline using distributed approaches

class nipype.pipeline.plugins.base.GraphPluginBase(plugin_args=None)

Bases: nipype.pipeline.plugins.base.PluginBase

Base class for plugins that distribute graphs to workflows

run(graph, config, updatehash=False)

Instruct the plugin to execute the workflow graph.

The core plugin member that should be implemented by all plugins.

Parameters:
  • graph – a networkx, flattened DAG to be executed

  • config (config) – a nipype.config object

  • updatehash (bool) – whether cached nodes with stale hash should be just updated.

class nipype.pipeline.plugins.base.PluginBase(plugin_args=None)

Bases: object

Base class for plugins.

run(graph, config, updatehash=False)

Instruct the plugin to execute the workflow graph.

The core plugin member that should be implemented by all plugins.

Parameters:
  • graph – a networkx, flattened DAG to be executed

  • config (config) – a nipype.config object

  • updatehash (bool) – whether cached nodes with stale hash should be just updated.

class nipype.pipeline.plugins.base.SGELikeBatchManagerBase(template, plugin_args=None)

Bases: nipype.pipeline.plugins.base.DistributedPluginBase

Execute workflow with SGE/OGE/PBS like batch system