de.unistuttgart.isa.liquidsvm

Class Config

java.lang.Object

de.unistuttgart.isa.liquidsvm.Config

All Implemented Interfaces:

Serializable

public class Config
extends Object
implements Serializable

Author:

Philipp Thomann

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
Config()
Config(Config other)

Method Summary

Modifier and Type	Method and Description
Config	adaptivityControl(int d) Activates adaptivity in the hyperparameter grid.
Config	display(int d) Sets the display level starting at 0 which means no information.
List<Map.Entry<String,String>>	entries() Creates a List that contains all key-value pairs in the order they were set.
Config	folds(int d)
String	get(String key)
Config	gridChoice(int d) Sets the hyperparameter grid.
boolean	has(String key)
boolean	isAutoSelect()
boolean	isAutoTrain()
Config	partitionChoice(int d) Sets the partition choice.
Config	randomSeed(int d)
String	remove(Object key)
Config	scale(boolean d)
Config	scenario(String scenario) Sets the scenario.
Config	select(boolean b) Specifies whether selecting should start automagically at the end of initializing the SVM after the automatic training.
Config	set(String key, String value) Sets the value for the key and remembers the order in which it was added.
Config	threads(int d) Sets the number of threads to use.
Config	train(boolean b) Specifies whether training should start automagically at the end of initializing the SVM.
Config	useCells()
Config	useCells(boolean b)
Config	voronoi(int d)

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

Config

public Config()

Config

public Config(Config other)

Method Detail

set

public Config set(String key,
                  String value)

Sets the value for the key and remembers the order in which it was added. This order is used in entries(). It returns this Config to enable method chaining. Many of the other methods use this.

Parameters:

key -

value -

Returns:

this config

See Also:

entries()

remove

public String remove(Object key)

get

public String get(String key)

has

public boolean has(String key)

entries

public List<Map.Entry<String,String>> entries()

Creates a List that contains all key-value pairs in the order they were set.

Returns:

a list containing all key-value pairs

See Also:

set(String, String)

scenario

public Config scenario(String scenario)

Sets the scenario.

Parameters:

scenario - can be any of "LS", "EX", "QT", "MC", "NPL", "ROC"

Returns:

this Config

See Also:

set(String, String), SVM.LS, SVM.MC

display

public Config display(int d)

Sets the display level starting at 0 which means no information.

Parameters:

d - display info level

Returns:

this Config

See Also:

set(String, String)

threads

public Config threads(int d)

Sets the number of threads to use. Default 0 means to use all physical cores, -1 means that all but one of the physical cores are used.s

Parameters:

d - number of threads

Returns:

this Config

See Also:

set(String, String)

partitionChoice

public Config partitionChoice(int d)

Sets the partition choice. The default 0 means no partitioning. See Cells.

Parameters:

d - display info level

Returns:

this Config

See Also:

set(String, String)

gridChoice

public Config gridChoice(int d)

Sets the hyperparameter grid. See Hyperparameter Grid.

Parameters:

d -

Returns:

this Config

See Also:

set(String, String)

adaptivityControl

public Config adaptivityControl(int d)

Activates adaptivity in the hyperparameter grid. See Adaptive Grid.

Parameters:

d -

Returns:

this Config

See Also:

set(String, String)

randomSeed

public Config randomSeed(int d)

folds

public Config folds(int d)

voronoi

public Config voronoi(int d)

scale

public Config scale(boolean d)

useCells

public Config useCells()

useCells

public Config useCells(boolean b)

train

public Config train(boolean b)

Specifies whether training should start automagically at the end of initializing the SVM. Default is true.

Parameters:

b -

Returns:

this Config

See Also:

SVM.train(String...)

select

public Config select(boolean b)

Specifies whether selecting should start automagically at the end of initializing the SVM after the automatic training. Default is true. This only has effect if #isAutoTrain() is true.

Parameters:

b -

Returns:

this Config

See Also:

SVM.select(String...)

isAutoTrain

public boolean isAutoTrain()

isAutoSelect

public boolean isAutoSelect()