all_train_splitter

Module: missions.nodes.splitter.all_train_splitter

Use all available data for training

Inheritance diagram for pySPACE.missions.nodes.splitter.all_train_splitter:

Inheritance diagram of pySPACE.missions.nodes.splitter.all_train_splitter

AllTrainSplitterNode

class pySPACE.missions.nodes.splitter.all_train_splitter.AllTrainSplitterNode(non_persistent=False, use_test_data=True, *args, **kwargs)[source]

Bases: pySPACE.missions.nodes.base_node.BaseNode

Use all available data for training

This node allows subsequent nodes to use all available labeled data for training. Accordingly, no data for testing is provided.

Parameters

Exemplary Call

-
    node : All_Train_Splitter
Author:

Jan Hendrik Metzen (jhm@informatik.uni-bremen.de)

Created:

2009/01/07

POSSIBLE NODE NAMES:
 
  • All_Train_Splitter
  • AllTrainSplitter
  • AllTrainSplitterNode
POSSIBLE INPUT TYPES:
 
  • PredictionVector
  • FeatureVector
  • TimeSeries

Class Components Summary

input_types
is_split_node() Returns whether this is a split node.
request_data_for_testing() Returns the data for testing of subsequent nodes
request_data_for_training(use_test_data) Returns the data for training of subsequent nodes
train_sweep(use_test_data) Performs the actual training of the node.
use_next_split() Use the next split of the data into training and test data.
__init__(non_persistent=False, use_test_data=True, *args, **kwargs)[source]
is_split_node()[source]

Returns whether this is a split node.

use_next_split()[source]

Use the next split of the data into training and test data.

Returns True if more splits are available, otherwise False.

This method is useful for benchmarking

train_sweep(use_test_data)[source]

Performs the actual training of the node.

Note

Split nodes cannot be trained

request_data_for_training(use_test_data)[source]

Returns the data for training of subsequent nodes

request_data_for_testing()[source]

Returns the data for testing of subsequent nodes

input_types = ['PredictionVector', 'FeatureVector', 'TimeSeries']