Let's assume I have a standard normal distribution $P(A)$ . We know $P(A = 1) = 0$, since cont. probability distribution have to be integrated over two values.
If I expend logic to joint multivariate gaussian $P(A=1,B=1)$, this as well have to be zero, correct?
But what if use multivariate conditional distributions $P(A|B=1)$ or $P(A=1|B)$ .Intution says to me, this would not be zero, that in both of this cases there are set of values, which could be expressed in analytical form... Would this be correct?