This class implements the separation of data used for analysis and analysis of the data. The latter can include computation of association rules as in showrules().
● Data Source:
CranContrib
● Keywords:
● Alias: PreProCombClass-class
●
0 images
|
prepro
(Package: preprocomb) :
the MAIN function for interactive use.
prepro() takes data, transforms it according to the given preprocessor and computes statistics of the transformed data. The main use case is the chaining of the preprocessors as show in the examples below.
● Data Source:
CranContrib
● Keywords:
● Alias: prepro
●
0 images
|
setpreprocessor
(Package: preprocomb) :
constructor function for adding a new preprocessing technique to the system
The main argument is the operation that is executed to transform the data such as "na.omit(basedata)" for removing rows that have missing values. An operation can process either only the numeric columns or also the class label column.
● Data Source:
CranContrib
● Keywords:
● Alias: setpreprocessor
●
0 images
|
initializedataclassobject() is a constructor function for creating a DataClass object. The main use case is adding of new preprocessing techniques to the framework by the user. Added preprocessing techniques (i.e. functions) take as input and must return a DataClass object. See setpreprocessor().
● Data Source:
CranContrib
● Keywords:
● Alias: initializedataclassobject
●
0 images
|
getprogrammaticprediction
(Package: preprocomb) :
parallel computation of classification accuracy holdout rounds
This function is used internally and exported for package 'metaheur'.
● Data Source:
CranContrib
● Keywords:
● Alias: getprogrammaticprediction
●
0 images
|
PreprocessorClass-class
(Package: preprocomb) :
an abstract class from which concrete preprocessor (sub) classes are inhereted.
Inheritance is controlled by setpreprocessor() function.
● Data Source:
CranContrib
● Keywords:
● Alias: PreprocessorClass-class
●
0 images
|
preprocomb
(Package: preprocomb) :
the MAIN function of programmatic use.
preprocomb executes the computation of classification accuracy, hopkins statistic and ORH outlier score. An alternative to preprocomb is to use package 'metaheur' for faster finding of near-optimal combinations.
● Data Source:
CranContrib
● Keywords:
● Alias: preprocomb
●
0 images
|
showrules
(Package: preprocomb) :
shows association rules for classification accuracy.
Classification accuracy label 'high' corresponds to best twenty percent and 'low' for the rest.
● Data Source:
CranContrib
● Keywords:
● Alias: showrules
●
0 images
|
transformdata is a generic preprocessing function. Its methods are defined by setpreprocessor(). The function is intented for package internal use, but exported so that classes can be inhereted from it.
● Data Source:
CranContrib
● Keywords:
● Alias: transformdata
●
0 images
|
setphase
(Package: preprocomb) :
constructor function for defining a preprocessing phase.
Preprocessing phases consist of preprocessing techniques defined with setpreprocessor(). Phases can be defined with setphase() and combined to a grid of combinations with setgrid().
● Data Source:
CranContrib
● Keywords:
● Alias: setphase
●
0 images
|