Free Statistics

TANAGRA

Tanagra is defined by the author (Ricco Rakotomalala, professor at the University in Lione, France) as a free data mining software for research and education. Tanagra is available only for Windows and it is downloadable freely. Tanagra is really an open source project and the author allows to download also the whole source code so to give to everyone the possibility of developing this software.
The interface of Tanagra consists of three parts which identify the way of working in the program:

1) The stream diagram (built in a tree structure) which represents the various steps of analysis you want to carry out in Tanagra.
2) The single operators (or components) each of those (identified in the diagram by a node) represents an operation done on the data: for example there is a component to visualize the dataset, another one to perform a regression analysis and so on. The set of these operators is placed at the bottom of the interface and it is subdivided in categories. Some of them correspond to the various types of analysis (descriptive statistics, multiple linear regression, clustering , principal component analysis, multiple correspondance analysis and so on), other ones are related to dataset manipulation (importation, visualization, attribute definition).
3) The report in html format subdivived in two parts:the description of the analysis parameters and the results.

Tanagra can import text file format with tab separator.
You can save the set of operations done on a dataset with two different formats:
- with .bdm format, by which you can save the chain of analyses built. This format can be opened only by Tanagra, so prefer to use it only if data should no longer change.
- with .tdm format, by which all the operations are saved in a file which can be opened by any text editor, so you can easily change later both the steps of analysis and/or the reference to the data source. On the other hand, the .tdm format is much slower to load than the .bdm one.
Beyond the various types of analysis above mentioned, Tanagra (conforming to its purpose to be useful in education too) can run several supervised learning process, for example the linear discriminant analysis and the binary logistic regression.
As tutorials there are (in the download section of the Tanagra website) some detailed examples in pdf format about the several methods of analysis you can perform and about the methods evaluation.
In conclusion Tanagra is a very useful software, which, thanks to its tree structure, allows to lead in parallel several concomitant analyses on the same data and to compare the results.
Among the few limitations of this software there are the impossibility to access to external database and the one to manage missing data.

Compare Tanagra with other free software Go to Tanagra website Back to Free Statistical Software

Home

Free Statistical Software

Free Math Software

Learning Statistics

Statistical Data

Java Statistics