Uninet is a standalone uncertainty analysis software package. Its main focus is dependence modelling for high dimensional distributions. Random variables can be coupled using a Bayesian network (Bayesnet), probability vines or dependence trees.
Download the latest installer: Download Page
Read Uninet help file describing the software in detail: UninetHelp.pdf (document size 2.1 MB)
Besides the Uninet application and its GUI, the functional core library UninetEngine can be used directly. The UninetEngine COM library is an extensive, object oriented, language-independent library containing over 70 classes, with over 500 methods (functions).
The library can be used from a wide variety of programming and scripting languages: C++, C#, VB.net, Delphi, Matlab, R, Octave and VBA (as used by e.g. Excel) are a few languages in which frameworks using UninetEngine have been or are being written.
There are a number of extra facilities accessible through the programmatic interface (e.g. a Bayesnet can be specified via a product-moment correlation matrix).
There are different Bayesnet samplers accessible through the programmatic interface (e.g. the pure memory sampler).
1. Random variables view – specify input random variables for your model and assign distributions.
2. Bayesnet view – build your model with probabilistic nodes, functional nodes and arcs.
3. Specify (conditional) rank correlation coefficients on the arcs.
4. Specify formulae for functional nodes.
5. Sample the model and view results.
6. Call satellite programs for further analysis of the sample.
7. Explore multivariate joint distributions with Unigraph.
8. Carry out sensitivity analysis with Unisens.
And many more features to discover!