org.encog.app.analyst.wizard
Class AnalystWizard

java.lang.Object
  extended by org.encog.app.analyst.wizard.AnalystWizard

public class AnalystWizard
extends Object

The Encog Analyst Wizard can be used to create Encog Analyst script files from a CSV file. This class is typically used by the Encog Workbench, but it can easily be used from any program to create a starting point for an Encog Analyst Script. Several items must be provided to the wizard. Desired Machine Learning Method: This is the machine learning method that you would like the wizard to use. This might be a neural network, SVM or other supported method. Normalization Range: This is the range that the data should be normalized into. Some machine learning methods perform better with different ranges. The two ranges supported by the wizard are -1 to 1 and 0 to 1. Goal: What are we trying to accomplish. Is this a classification, regression or autoassociation problem.


Field Summary
static int DEFAULT_EVAL_PERCENT
          The default evaluation percent.
static double DEFAULT_TRAIN_ERROR
          The default training error.
static int DEFAULT_TRAIN_PERCENT
          The default training percent.
static String FILE_BALANCE
          The balanced file.
static String FILE_CLUSTER
          The clustered file.
static String FILE_CODE
          The generated code file.
static String FILE_EVAL
          The evaluation file.
static String FILE_EVAL_NORM
          The eval file normalization file.
static String FILE_ML
          The machine learning file.
static String FILE_NORMALIZE
          The normalized file.
static String FILE_OUTPUT
          The output file.
static String FILE_PRE
          The processed data.
static String FILE_RANDOM
          The randomized file.
static String FILE_RAW
          The raw file.
static String FILE_TRAIN
          The training file.
static String FILE_TRAINSET
          The training set.
 
Constructor Summary
AnalystWizard(EncogAnalyst theAnalyst)
          Construct the analyst wizard.
 
Method Summary
 TargetLanguage getCodeTargetLanguage()
           
 int getEvidenceSegements()
           
 AnalystGoal getGoal()
           
 int getLagWindowSize()
           
 int getLeadWindowSize()
           
 double getMaxError()
           
 WizardMethodType getMethodType()
           
 HandleMissingValues getMissing()
           
 NormalizeRange getRange()
           
 AnalystField getTargetField()
           
 boolean isCodeEmbedData()
           
 boolean isIncludeTargetField()
           
 boolean isNaiveBayes()
           
 boolean isPreprocess()
           
 boolean isTaskBalance()
           
 boolean isTaskCluster()
           
 boolean isTaskNormalize()
           
 boolean isTaskRandomize()
           
 boolean isTaskSegregate()
           
 void reanalyze()
          Reanalyze column ranges.
 void setCodeEmbedData(boolean codeEmbedData)
           
 void setCodeTargetLanguage(TargetLanguage codeTargetLanguage)
           
 void setEvidenceSegements(int evidenceSegements)
           
 void setGoal(AnalystGoal theGoal)
          Set the goal.
 void setIncludeTargetField(boolean theIncludeTargetField)
           
 void setLagWindowSize(int theLagWindowSize)
           
 void setLeadWindowSize(int theLeadWindowSize)
           
 void setMaxError(double maxError)
           
 void setMethodType(WizardMethodType theMethodType)
           
 void setMissing(HandleMissingValues missing)
           
 void setNaiveBayes(boolean naiveBayes)
           
 void setPreprocess(boolean preprocess)
           
 void setRange(NormalizeRange theRange)
           
 void setTargetField(AnalystField theTargetField)
          Set the target field.
 void setTargetField(String theTargetField)
           
 void setTaskBalance(boolean theTaskBalance)
           
 void setTaskCluster(boolean theTaskCluster)
           
 void setTaskNormalize(boolean theTaskNormalize)
           
 void setTaskRandomize(boolean theTaskRandomize)
           
 void setTaskSegregate(boolean theTaskSegregate)
           
 void wizard(File analyzeFile, boolean b, AnalystFileFormat format)
          Analyze a file.
 void wizard(URL url, File saveFile, File analyzeFile, boolean b, AnalystFileFormat format)
          Analyze a file at the specified URL.
 void wizardRealTime(List<SourceElement> sourceData, File csvFile, int backwardWindow, int forwardWindow, PredictionType prediction, String predictField)
           
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Field Detail

DEFAULT_TRAIN_PERCENT

public static final int DEFAULT_TRAIN_PERCENT
The default training percent.

See Also:
Constant Field Values

DEFAULT_EVAL_PERCENT

public static final int DEFAULT_EVAL_PERCENT
The default evaluation percent.

See Also:
Constant Field Values

DEFAULT_TRAIN_ERROR

public static final double DEFAULT_TRAIN_ERROR
The default training error.

See Also:
Constant Field Values

FILE_PRE

public static final String FILE_PRE
The processed data.

See Also:
Constant Field Values

FILE_RAW

public static final String FILE_RAW
The raw file.

See Also:
Constant Field Values

FILE_NORMALIZE

public static final String FILE_NORMALIZE
The normalized file.

See Also:
Constant Field Values

FILE_RANDOM

public static final String FILE_RANDOM
The randomized file.

See Also:
Constant Field Values

FILE_TRAIN

public static final String FILE_TRAIN
The training file.

See Also:
Constant Field Values

FILE_EVAL

public static final String FILE_EVAL
The evaluation file.

See Also:
Constant Field Values

FILE_EVAL_NORM

public static final String FILE_EVAL_NORM
The eval file normalization file.

See Also:
Constant Field Values

FILE_TRAINSET

public static final String FILE_TRAINSET
The training set.

See Also:
Constant Field Values

FILE_ML

public static final String FILE_ML
The machine learning file.

See Also:
Constant Field Values

FILE_OUTPUT

public static final String FILE_OUTPUT
The output file.

See Also:
Constant Field Values

FILE_BALANCE

public static final String FILE_BALANCE
The balanced file.

See Also:
Constant Field Values

FILE_CLUSTER

public static final String FILE_CLUSTER
The clustered file.

See Also:
Constant Field Values

FILE_CODE

public static final String FILE_CODE
The generated code file.

See Also:
Constant Field Values
Constructor Detail

AnalystWizard

public AnalystWizard(EncogAnalyst theAnalyst)
Construct the analyst wizard.

Parameters:
theAnalyst - The analyst to use.
Method Detail

getGoal

public AnalystGoal getGoal()
Returns:
The analyst goal.

getLagWindowSize

public int getLagWindowSize()
Returns:
the lagWindowSize

getLeadWindowSize

public int getLeadWindowSize()
Returns:
the leadWindowSize

getMethodType

public WizardMethodType getMethodType()
Returns:
the methodType

getRange

public NormalizeRange getRange()
Returns:
the range

getTargetField

public AnalystField getTargetField()
Returns:
Get the target field.

isIncludeTargetField

public boolean isIncludeTargetField()
Returns:
the includeTargetField

isTaskBalance

public boolean isTaskBalance()
Returns:
the taskBalance

isTaskCluster

public boolean isTaskCluster()
Returns:
the taskCluster

isTaskNormalize

public boolean isTaskNormalize()
Returns:
the taskNormalize

isTaskRandomize

public boolean isTaskRandomize()
Returns:
the taskRandomize

isTaskSegregate

public boolean isTaskSegregate()
Returns:
the taskSegregate

reanalyze

public void reanalyze()
Reanalyze column ranges.


setGoal

public void setGoal(AnalystGoal theGoal)
Set the goal.

Parameters:
theGoal - The goal.

setIncludeTargetField

public void setIncludeTargetField(boolean theIncludeTargetField)
Parameters:
theIncludeTargetField - the includeTargetField to set

setLagWindowSize

public void setLagWindowSize(int theLagWindowSize)
Parameters:
theLagWindowSize - the lagWindowSize to set

setLeadWindowSize

public void setLeadWindowSize(int theLeadWindowSize)
Parameters:
theLeadWindowSize - the leadWindowSize to set

setMethodType

public void setMethodType(WizardMethodType theMethodType)
Parameters:
theMethodType - the methodType to set

setRange

public void setRange(NormalizeRange theRange)
Parameters:
theRange - the range to set

setTargetField

public void setTargetField(AnalystField theTargetField)
Set the target field.

Parameters:
theTargetField - The target field.

setTaskBalance

public void setTaskBalance(boolean theTaskBalance)
Parameters:
theTaskBalance - the taskBalance to set

setTaskCluster

public void setTaskCluster(boolean theTaskCluster)
Parameters:
theTaskCluster - the taskCluster to set

setTaskNormalize

public void setTaskNormalize(boolean theTaskNormalize)
Parameters:
theTaskNormalize - the taskNormalize to set

setTaskRandomize

public void setTaskRandomize(boolean theTaskRandomize)
Parameters:
theTaskRandomize - the taskRandomize to set

setTaskSegregate

public void setTaskSegregate(boolean theTaskSegregate)
Parameters:
theTaskSegregate - the taskSegregate to set

wizardRealTime

public void wizardRealTime(List<SourceElement> sourceData,
                           File csvFile,
                           int backwardWindow,
                           int forwardWindow,
                           PredictionType prediction,
                           String predictField)

wizard

public void wizard(File analyzeFile,
                   boolean b,
                   AnalystFileFormat format)
Analyze a file.

Parameters:
analyzeFile - The file to analyze.
b - True if there are headers.
format - The file format.

wizard

public void wizard(URL url,
                   File saveFile,
                   File analyzeFile,
                   boolean b,
                   AnalystFileFormat format)
Analyze a file at the specified URL.

Parameters:
url - The URL to analyze.
saveFile - The save file.
analyzeFile - The Encog analyst file.
b - True if there are headers.
format - The file format.

getMissing

public HandleMissingValues getMissing()
Returns:
the missing

setMissing

public void setMissing(HandleMissingValues missing)
Parameters:
missing - the missing to set

isNaiveBayes

public boolean isNaiveBayes()
Returns:
the naiveBayes

setNaiveBayes

public void setNaiveBayes(boolean naiveBayes)
Parameters:
naiveBayes - the naiveBayes to set

getEvidenceSegements

public int getEvidenceSegements()
Returns:
the evidenceSegements

setEvidenceSegements

public void setEvidenceSegements(int evidenceSegements)
Parameters:
evidenceSegements - the evidenceSegements to set

getMaxError

public double getMaxError()

setMaxError

public void setMaxError(double maxError)

setTargetField

public void setTargetField(String theTargetField)

getCodeTargetLanguage

public TargetLanguage getCodeTargetLanguage()
Returns:
the codeTargetLanguage

setCodeTargetLanguage

public void setCodeTargetLanguage(TargetLanguage codeTargetLanguage)
Parameters:
codeTargetLanguage - the codeTargetLanguage to set

isCodeEmbedData

public boolean isCodeEmbedData()
Returns:
the codeEmbedData

setCodeEmbedData

public void setCodeEmbedData(boolean codeEmbedData)
Parameters:
codeEmbedData - the codeEmbedData to set

isPreprocess

public boolean isPreprocess()

setPreprocess

public void setPreprocess(boolean preprocess)


Copyright © 2014. All Rights Reserved.