Inference

PPL can check coherency and perform extensions (inferences) in a knowledge base. To do so, first you have to insert some assessments (section 5.4) and then, if they are coherent, PPL will computate the upper and lower probabilities for any adicional event.

After inserting the assessments, you have two choices: just check their coherency and/or extend them. If you choose the first option, the function to be used is the PPL.checkCoherence:

»> PPL.checkCoherence()

The result of this interaction should be `coherent!' or `incoherent!'. In case you want to perform inference:

»> s2 = `A$\wedge$C'

»> PPL.extension(s2)

Where s2 is the adicional assessment that you want to calculate the upper and lower probabilities. If the assessments are coherent, the result of this interaction should be something like this:

Lower: 0.2

Upper: 0.6

When the function extension is used, PPL calls the function checkCoherence and if it returns `coherent!', then the probabilities interval is calculated.