weka.gui.explorer.Explorer.props Maven / Gradle / Ivy

Go to download
Show more of this group Show more artifacts with this name
Show all versions of weka-dev Show documentation
The Waikato Environment for Knowledge Analysis (WEKA), a machine learning workbench. This version represents the developer version, the "bleeding edge" of development, you could say. New functionality gets added to this version.
There is a newer version: 3.9.6
Show newest version
# This props file contains default values for the Weka Explorer.
#
# Notes:
# - backslashes within options, e.g., for the default "Classifier", need
#   to be doubled (the backslashes get interpreted already when a property
#   is read).
#
# Author  FracPete (fracpete at waikato dot ac dot nz)
# Version $Revision: 7526 $

# if set to true the Capabilities filters in the GOE will be initialized
# based on the full dataset that has been loaded into the Explorer 
# otherwise only the header (true|false)
# Note: The tabs in the Explorer have their own class combobox, which means 
#       that the data has to be inspected several times (changing the class
#       combobox only leads to an inspection of the data in the current tab), 
#       which can be slow on big datasets.
InitGenericObjectEditorFilter=True

# The tabs to display apart from the PreprocessPanel.
#
# The classes listed here must import the weka.gui.explorer.Explorer.ExplorerPanel
# interface. Optionally, they can also import the 
# weka.gui.explorer.Explorer.LogHandler interface if they want to use the logging
# functionality of the Explorer and the 
# weka.gui.exporer.Explorer.CapabilitiesFilterChangeListener interface 
# in case they need to know when the Capabilities have changed, e.g., when a 
# new dataset has been loaded into the Explorer.
#
# Additional options follow the classname after a colon.
# Currently supported options are:
#   standalone - the tab does not depend on the PreprocessPanel to load the data first
#
Tabs=weka.gui.explorer.ClassifierPanel,\
     weka.gui.explorer.ClustererPanel,\
     weka.gui.explorer.AssociationsPanel,\
     weka.gui.explorer.AttributeSelectionPanel,\
     weka.gui.explorer.VisualizePanel

# the initial directory for opening datasets.
# the following placeholders are recognized
#   %t - the temp directory
#   %h - the user's home directory
#   %c - the current directory
#   %% - gets replaced by a single percentage sign
InitialDirectory=%

# whether to create undo points
enableUndo=true

# the directory in which to create undo files
# %t is a placeholder for the standard system temp directory
undoDirectory=%t

# the default filter, including options (can be left empty)
Filter=

# the default classifier in the classify tab, including options
# (default is ZeroR)
Classifier=weka.classifiers.rules.ZeroR

# the default test mode in the classify tab 
# (according to "testMode" variable in startClassifier method)
# 1 - cross-validation
# 2 - percentage split
# 3 - use training set
# 4 - supplied test set
# (default is 1 - CV)
ClassifierTestMode=1

# the default number of folds for CV in the classify tab 
# (default is 10)
ClassifierCrossvalidationFolds=10

# the default percentage split % in the classify tab (integer: 1-99)
# (default is 66)
ClassifierPercentageSplit=66

# whether the classifier model is output (true|false)
# (default is true)
ClassifierOutputModel=true

# whether additional per-class stats of the classifier model are 
# output (true|false)
# (default is true)
ClassifierOutputPerClassStats=true

# whether the entropy based evaluation measures of the classifier model are 
# output (true|false)
# (default is false)
ClassifierOutputEntropyEvalMeasures=false

# whether the confusion matrix is output for the classifier (true|false)
# (default is true)
ClassifierOutputConfusionMatrix=true

# whether the predictions of the classifier are stored for visulization 
# purposes (true|false)
# (default is true)
ClassifierStorePredictionsForVis=true

# whether the predictions of the classifier output as well (true|false)
# (default is false)
ClassifierOutputPredictions=false

# lists the attributes indices to output in addition to the predictions
# (default is "")
ClassifierOutputAdditionalAttributes=

# whether the evaluation of the classifier is done cost-sensitively (true|false)
# (default is false)
# Note: a cost matrix still has to be provided!
ClassifierCostSensitiveEval=false

# the default random seed in the classify tab
# (default is 1)
ClassifierRandomSeed=1

# whether the order is preserved in case of percentage split in the classifier 
# tab 
# (default is false)
ClassifierPreserveOrder=false

# whether the source code of a Sourcable classifier is output as well in the 
# classifier tab 
# (default is false)
ClassifierOutputSourceCode=false

# the default classname of a Sourcable classifier in the classifier tab 
# (default is Foobar)
ClassifierSourceCodeClass=WekaClassifier

# the class (incl. options) for collecting the predictions and turning them 
# into plotable instances for displaying the classifier errors.
ClassifierErrorsPlotInstances=weka.gui.explorer.ClassifierErrorsPlotInstances

# The minimum plot size for numeric attributes (when visualizing classifier errors)
ClassifierErrorsMinimumPlotSizeNumeric=1

# The maximum plot size for numeric attributes (when visualizing classifier errors)
ClassifierErrorsMaximumPlotSizeNumeric=20

# the default clusterer, including options
# (default is EM)
Clusterer=weka.clusterers.EM

# the default test mode in the cluster tab 
# (according to "testMode" variable in startClusterer method)
# 2 - percentage split
# 3 - use training set
# 4 - supplied test set
# 5 - classes to clusters evaluation
# (default is 3 - training set)
ClustererTestMode=3

# whether the clusters are stored for visualization purposes (true|false)
# (default is true)
ClustererStoreClustersForVis=true

# the class (incl. options) for collecting the predictions and turning them 
# into plotable instances for displaying the cluster assignments.
ClustererAssignmentsPlotInstances=weka.gui.explorer.ClustererAssignmentsPlotInstances

# the default associator, including options
# (default is Apriori)
Associator=weka.associations.Apriori

# the default attribute evaluator, including options
# (default is CfsSubsetEval)
ASEvaluation=weka.attributeSelection.CfsSubsetEval

# the default attribute selection search scheme, including options
# (default is BestFirst)
ASSearch=weka.attributeSelection.BestFirst

# the default test mode in the attribute selection tab 
# (according to "testMode" variable in startAttributeSelection method)
# 0 - use full training set
# 1 - cross-validation
# (default is 0 - full training set)
ASTestMode=0

# the default number of folds for CV in the attribute selection tab 
# (default is 10)
ASCrossvalidationFolds=10

# the default random seed in the attribute selection tab
# (default is 1)
ASRandomSeed=1