ZDDs and FDDs make use of the interpretation of decision diagrams as sets of sets. While ZDDs store the full minterms of a DNF as sets of sets, FDDs store the monomials of a RMNF as sets of sets. How a set of set is represented by a decision diagram is explained on slides 140-142 of the BDD chapter of VRS.

For example, consider the following sets of sets {{a,b,c},{a,c},{b,c},{}} that represents the monomials of a Reed-Muller normal form. Converting this into a decision diagram with variable ordering c<b<a works as follows: First, we partition the sets into those containing "a", and those not containing "a":

- {a,b,c},{a,c} which yield {b,c},{c} for the positive edge
- {b,c},{} for the negative edge

For the positive edge, we continue with paritioning by "b":

- {b,c} which yield {c} for the positive edge
- {c} for the negative edge

So, we get a node labelled with "b" having two edges for the set representation of {c} which is just a node labelled with c having positive and negative edges to the 1-leaf and 0-leaf, respectively.

For the negative edge of "a", we also continue with paritioning by "b":

- {b,c} which yield {c} for the positive edge
- {} for the negative edge

So, we get a node labelled with "b" having a positive edge to the c-node that we already have, and a negative edge to the 1-leaf. Note that the 0-leaf is represented by {} and the 1-leaf is represented by {{}}.