I was going through the Probabilistic Graphical Modelling course by Stanford and they used a network such as this one-https://imgur.com/gallery/k0C8FY2
Now if we want to sample P(A|B), how would we do that? My understanding is that we need to sample A, C and D while keeping the assigned value of B constant. However, the factors only provide us with Joint Probability table of P(A,B,C,D), so how would we compute P(C|B,D)? I am new to Markov Networks and Gibbs sampling so let me know where I might have gone wrong.
EDIT-How can we sample them if we don't know the individual probabilities or the conditional probabilities?