VIBES

Variational Inference for Bayesian Networks

SourceForge.net Logo
[Sourceforge project page]
John Winn, January 2004
Overview Tutorial Examples Help
previous    Contents   next

2. Loading the Matlab data into VIBES

Before we can load the data, we must create a model to attach the data to. This is achieved by creating a node with the name x (this must be the same name as the data matrix in the Matlab .mat file). As the data matrix is two-dimensional, the node is placed inside two plates N and d (again corresponding to the names of the variables in the matlab file). The model then looks like the diagram below. You must now save the model to the directory where the Matlab file is stored.

Note: If you want to skip constructing this network by hand, it is in the tutorial file called DataOnly.xml.

We are now ready to load the Matlab file. The filename (in this case MixGaussianData2D.mat) is entered as shown in the diagram (if necessary clicking to unselect any selected nodes or plates). Choosing File->Load data will then load the data into the node and also set the size of the N and d plates. The node x will then be marked as observed (shown with a bold edge) and the observed data can be inspected by double-clicking the node with the mouse: