On set with $n$ elements, number of equivalence relations is greater then the number of partial-order relations? Or the opposite? Transitive relations make it hard to count these numbers. Can anyone help?
2026-03-25 12:50:11.1774443011
On
On set with $n$ elements, number of equivalence relations is greater then the number of partial-order relations?
89 Views Asked by Bumbble Comm https://math.techqa.club/user/bumbble-comm/detail At
2
There are 2 best solutions below
0
On
The Wikipedia link here has the answer: https://en.wikipedia.org/wiki/Partially_ordered_set#Number_of_partial_orders
It appears that there are more partial orders than equivalence relations for n=2 through n=19. It also seems likely that this pattern continues, but as no distinct formulas are given for them, this is not guaranteed.
The partial orders are more numerous than the equivalence relations. First, let's order the elements of the set. Indeed, without loss of generality, we'll make the set $S = \lbrace 1, 2, \ldots, n\rbrace$.
Recall that equivalence relations correspond to partitions, i.e. sets $\lbrace P_1, \ldots, P_k \rbrace$ such that $\bigcup_{i=1}^k P_k = S$, $P_i \neq \emptyset$ for all $i$, and $P_i \cap P_j \neq \emptyset \implies i = j$. Let's fix such a partition. Moreover, we can totally order the parts in the partition by forcing $P_i \le P_j \iff \min P_i \le \min P_j$.
Then, define a partial order on $S$ by $a \preceq b$ if and only if $a \in P_i$ and $b \in P_j$ such that $P_i \le P_j$. It's not difficult to show reflexivity, anti-symmetry, and transitivity of this relation. Basically, I'm constructing the partial order through its Hasse Diagram: the partitions form the layers of the diagram, and arrows pass freely up the diagram between the layers.
So, we have a method for constructing a partial order from an equivalence relation. This forms a function from the equivalence relations on $S$ to the partial orders on $S$. If we can show that such a function is injective, we are done; the equivalence relations are equinumerous with a subset of the partial orders.
Suppose we have two equivalence relations, with corresponding partitions $\lbrace P_1, \ldots, P_k \rbrace$ and $\lbrace Q_1, \ldots, Q_m\rbrace$, such that the above method produces the same equivalence relation. Note that, according to the preceding construction, $k$ and $m$ are both the size of the longest chain (totally ordered subset) of the poset, so $k = m$. Moreover, the parts of the partition correspond to the maximal anti-chains (sets that have no relationship between any two elements), so the partitions are the same.
Therefore, yes, the partial orders are more numerous than the equivalence relations.