Orange

Orange Modules

Orange modules are intended to extend basic Orange's functionality, or provide wrappers for easier use of some data mining techniques. They are already included in normal Orange distribution. Following set of pages provides description, demos and examples for selected modules.

Note: The documentation posted on the web is updated from the CVS in real-time and refers to the latest snapshot of Orange. If you encounter any inconsistencies please compare the standalone documentation with the one on the web.

orngAssoc
A few things for association rules.
orngBayes
Wrapper around Orange's naive Bayesian learner that makes it easier to use m-estimation; it can also print out the model.
orngCA
Correspondence analysis.
orngC45
A module with a function that prints out C4.5 trees in exactly the same format as Quinlan's C4.5.
orngCI
Constructive induction (function decomposition methods, HINT, Kramer's constructive induction method).
orngClustering
Various clustering methods and associated miscellaneous utilities.
orngCN2
A set of classes and functions for learning rules (based on CN2).
orngDisc
Wrapper around Orange's categorization techniques for continuous attributes.
orngEnsemble
Bagging, boosting, random forests.
orngEval
Obsolete, included for compatibility with past version of Orange. Use orngTest and orngStat instead.
orngFSS
Feature subset selection.
orngImpute
Imputation wrappers for learners and classifiers.
orngLinProj
Implements the FreeViz method by Demsar et al.
orngLookup
Functions for working with classifiers with stored tables of examples.
orngLR
Wrappers for easier use of Orange's classes for logistic regression
orngMDS
Multidimensional scaling.
orngMisc
Miscellaneous functions, including various counters and selections of optimal objects in a sequence.
orngMySQL
Interface to MySQL.
orngNetwork
Network analysis and layout optimization.
orngOutlier
Outlier detection.
orngReinforcement
Reinforcement learning.
orngServerFiles
Orange's file repository.
orngSQL
A new interface to any PEP 249 compliant RDBS. Supports both MySQL and Postgres.
orngStat
Computation of various statistics such as accuracy, sensitivity, specificity, and area under ROC from the experimental data from module orngTest.
orngSVM
Support vector machines.
orngTest
Data sampling and testing of learners, for instance cross-validation, leave-one out, random sampling, etc. The results can be used by orngStat to compute various statistics.
orngTree
Wrapper around Orange's classification tree inducer. Most notably, implements tree printout and visualization.
orngVizRank
Implementation of VizRank intelligent visualization method.
orngWrap
Classes for tuning arguments using internal cross-validation and for searching for threshold for optimal classification accuracy.

Additional modules for data mining (clustering, SVM wrapper for probability estimation, logistic regression, density estimation...) have been prepared by Aleks Jakulin and are accessible via his pages.