The previous framework can be extended to deal with continuous data, and a mixture of continuous and discrete data. From an operational perspective, nothing changes much with respect to the methods that we have discussed in the previous recipe. It is worth noting that bnlearn requires that continuous data be gaussian, which might not work in some cases.
Continuous and hybrid Bayesian networks via bnlearn
Getting ready
The bnlearn package needs to be installed using install.packages("bnlearn").
How to do it...
In the following example, we will work with the abalone...
