TheoremBase

Sketch: exchange the order of summation over pairs (S, x) with x in ASA_S; the contribution of each x is an alternating binomial sum, which is 1 if x lies in no AiA_i and 0 otherwise.

Proof

Write U=⋃i=1mAiU=\bigcup_{i=1}^{m}A_{i} and Y=Xβˆ–UY=X\setminus U. Throughout, UU, YY, the sets ASA_{S} and the sets T(x)T(x) introduced below are subsets of the finite sets XX and [m][m], hence finite (Finite Sets: the Pigeonhole Principle, Uniqueness of the Length, Subsets, Unions, Products, Images, Bounded Sets of Natural Numbers, Extreme Elements, Sets of Maps, Finite Unions and Finite Choice (iv)); P([m])\mathcal{P}([m]) is finite by Counting Subsets (ii); and S↦#ASS\mapsto\#A_{S} and S↦#SS\mapsto\#S are maps on P([m])\mathcal{P}([m]) (The Maps i ↦ #C(i) and S ↦ #S (i), The Maps i ↦ #C(i) and S ↦ #S (ii)); and cardinalities are read in R\mathbb{R}, where their sums keep their values (The Image of the Natural Numbers with Zero in a Commutative Ring Respects Zero, One, Sums, Products, Differences, Powers, and Finite Sums and Products (ii)).

Step 1 (which sets ASA_{S} contain a given element). For x∈Xx\in X let T(x)={i∈[m]:x∈Ai}T(x)=\{i\in[m]:x\in A_{i}\}. We claim that for every SβŠ†[m]S\subseteq[m],

x∈ASβ€…β€ŠβŸΊβ€…β€ŠSβŠ†T(x).x\in A_{S}\iff S\subseteq T(x).

For S=βˆ…S=\emptyset both sides hold, because Aβˆ…=XA_{\emptyset}=X. For nonempty SS both sides say that x∈Aix\in A_{i} for every i∈Si\in S. Moreover, by the definition of the union, T(x)=βˆ…T(x)=\emptyset if and only if xx lies in no AiA_{i}, that is, if and only if x∈Yx\in Y.

Step 2 (each term as a sum over XX). For S∈P([m])S\in\mathcal{P}([m]) and x∈Xx\in X put

e(S,x)={(βˆ’1)#S,x∈AS,0,xβˆ‰AS.e(S,x)=\begin{cases}(-1)^{\#S}, & x\in A_{S},\\ 0, & x\notin A_{S}.\end{cases}

For fixed SS the map e(S,β‹…)e(S,\cdot) vanishes off ASA_{S} and is constant on ASA_{S}, so by Sums over Finite Sets in a Commutative Ring and in an Ordered Field: Distributivity, Products of Sums, Vanishing Terms, Sums over Pairs, Expanding Products of Sums, Counting, Comparison, Monotonicity and the Triangle Inequality (iv) and Sums over Finite Sets in a Commutative Ring and in an Ordered Field: Distributivity, Products of Sums, Vanishing Terms, Sums over Pairs, Expanding Products of Sums, Counting, Comparison, Monotonicity and the Triangle Inequality (v)

βˆ‘x∈Xe(S,x)=βˆ‘x∈AS(βˆ’1)#S=(βˆ’1)#S #AS.\sum_{x\in X}e(S,x)=\sum_{x\in A_{S}}(-1)^{\#S}=(-1)^{\#S}\,\#A_{S}.

Step 3 (exchanging the order of summation). Summing Step 2 over SS and exchanging the two sums over P([m])Γ—X\mathcal{P}([m])\times X by Iterated Operations over Finite Sets: Singletons, Disjoint Unions, Reindexing, Products of Sets, Termwise Combination, Homomorphisms and Intervals (iv), applied to the addition of R\mathbb{R}, we get

βˆ‘S∈P([m])(βˆ’1)#S #AS=βˆ‘S∈P([m])Β βˆ‘x∈Xe(S,x)=βˆ‘x∈XΒ βˆ‘S∈P([m])e(S,x).\sum_{S\in\mathcal{P}([m])}(-1)^{\#S}\,\#A_{S}=\sum_{S\in\mathcal{P}([m])}\ \sum_{x\in X}e(S,x)=\sum_{x\in X}\ \sum_{S\in\mathcal{P}([m])}e(S,x).

Fix x∈Xx\in X. By Step 1, e(S,x)=(βˆ’1)#Se(S,x)=(-1)^{\#S} for S∈P(T(x))S\in\mathcal{P}(T(x)) and e(S,x)=0e(S,x)=0 for the other S∈P([m])S\in\mathcal{P}([m]); since P(T(x))βŠ†P([m])\mathcal{P}(T(x))\subseteq\mathcal{P}([m]), Sums over Finite Sets in a Commutative Ring and in an Ordered Field: Distributivity, Products of Sums, Vanishing Terms, Sums over Pairs, Expanding Products of Sums, Counting, Comparison, Monotonicity and the Triangle Inequality (iv) gives

βˆ‘S∈P([m])e(S,x)=βˆ‘S∈P(T(x))(βˆ’1)#S.\sum_{S\in\mathcal{P}([m])}e(S,x)=\sum_{S\in\mathcal{P}(T(x))}(-1)^{\#S}.

Step 4 (the contribution of one element). Let t=#T(x)t=\#T(x). If t=0t=0, then T(x)=βˆ…T(x)=\emptyset by Counting: Intervals, Empty Sets and Singletons, Injective Images, Subsets, Unions and Products (ii), its only subset is βˆ…\emptyset, and the sum has the single term (βˆ’1)#βˆ…=(βˆ’1)0=1(-1)^{\#\emptyset}=(-1)^{0}=1 (Iterated Operations over Finite Sets: Singletons, Disjoint Unions, Reindexing, Products of Sets, Termwise Combination, Homomorphisms and Intervals (i), Counting: Intervals, Empty Sets and Singletons, Injective Images, Subsets, Unions and Products (ii), Powers with Exponents in the Natural Numbers with Zero (ii)). If tβ‰ 0t\neq0, then 1≀t1\le t, and t=n+1t=n+1 for some n∈N0n\in\mathbb{N}_{0} (Arithmetic of Addition on Omega: Recursion Rules, Associativity, Commutativity, Cancellation and Compatibility with the Order (vii)). Then Counting Subsets (iii), with A=T(x)A=T(x) and g:N0β†’Rg:\mathbb{N}_{0}\to\mathbb{R}, g(j)=(βˆ’1)jg(j)=(-1)^{j}, and Binomial Identities (ii) give

βˆ‘S∈P(T(x))(βˆ’1)#S=βˆ‘j=0n+1(βˆ’1)j(n+1j)=0.\sum_{S\in\mathcal{P}(T(x))}(-1)^{\#S}=\sum_{j=0}^{n+1}(-1)^{j}\binom{n+1}{j}=0.

By Step 1 and Counting: Intervals, Empty Sets and Singletons, Injective Images, Subsets, Unions and Products (ii), t=0t=0 exactly when x∈Yx\in Y. So the inner sum in Step 3 is 11 for x∈Yx\in Y and 00 for x∈Xβˆ–Yx\in X\setminus Y.

Step 5 ((i)). By Steps 3 and 4, together with Sums over Finite Sets in a Commutative Ring and in an Ordered Field: Distributivity, Products of Sums, Vanishing Terms, Sums over Pairs, Expanding Products of Sums, Counting, Comparison, Monotonicity and the Triangle Inequality (iv) and Sums over Finite Sets in a Commutative Ring and in an Ordered Field: Distributivity, Products of Sums, Vanishing Terms, Sums over Pairs, Expanding Products of Sums, Counting, Comparison, Monotonicity and the Triangle Inequality (v),

βˆ‘S∈P([m])(βˆ’1)#S #AS=βˆ‘x∈Y1=#Y=#(Xβˆ–β‹ƒi=1mAi).\sum_{S\in\mathcal{P}([m])}(-1)^{\#S}\,\#A_{S}=\sum_{x\in Y}1=\#Y=\#\Big(X\setminus\bigcup_{i=1}^{m}A_{i}\Big).

Step 6 ((ii)). The set P([m])\mathcal{P}([m]) is the disjoint union of {βˆ…}\{\emptyset\} and P([m])βˆ–{βˆ…}\mathcal{P}([m])\setminus\{\emptyset\}, and the term at S=βˆ…S=\emptyset is (βˆ’1)0 #Aβˆ…=#X(-1)^{0}\,\#A_{\emptyset}=\#X. So by Iterated Operations over Finite Sets: Singletons, Disjoint Unions, Reindexing, Products of Sets, Termwise Combination, Homomorphisms and Intervals (ii) and Iterated Operations over Finite Sets: Singletons, Disjoint Unions, Reindexing, Products of Sets, Termwise Combination, Homomorphisms and Intervals (i), Step 5 reads

#Y=#X+βˆ‘S∈P([m])βˆ–{βˆ…}(βˆ’1)#S #AS.\#Y=\#X+\sum_{S\in\mathcal{P}([m])\setminus\{\emptyset\}}(-1)^{\#S}\,\#A_{S}.

Since XX is the disjoint union of UU and YY, #U+#Y=#X\#U+\#Y=\#X by Counting: Intervals, Empty Sets and Singletons, Injective Images, Subsets, Unions and Products (v). For every SS, (βˆ’1)#S+1=(βˆ’1)#S(βˆ’1)1=βˆ’(βˆ’1)#S(-1)^{\#S+1}=(-1)^{\#S}(-1)^{1}=-(-1)^{\#S} by Powers in a Commutative Ring, a Field and an Ordered Field: Exponent Laws, Factorisation, Geometric Sums, Monotonicity and Bernoulli's Inequality (i) and Powers in a Commutative Ring, a Field and an Ordered Field: Exponent Laws, Factorisation, Geometric Sums, Monotonicity and Bernoulli's Inequality (ii). Hence, multiplying the sum by βˆ’1-1 term by term (Sums over Finite Sets in a Commutative Ring and in an Ordered Field: Distributivity, Products of Sums, Vanishing Terms, Sums over Pairs, Expanding Products of Sums, Counting, Comparison, Monotonicity and the Triangle Inequality (i)),

#U=#Xβˆ’#Y=βˆ’βˆ‘S∈P([m])βˆ–{βˆ…}(βˆ’1)#S #AS=βˆ‘S∈P([m])βˆ–{βˆ…}(βˆ’1)#S+1 #AS.\#U=\#X-\#Y=-\sum_{S\in\mathcal{P}([m])\setminus\{\emptyset\}}(-1)^{\#S}\,\#A_{S}=\sum_{S\in\mathcal{P}([m])\setminus\{\emptyset\}}(-1)^{\#S+1}\,\#A_{S}.

Citations

Loading…

Dependencies

Uses0

Loading…

Comments

Log in to comment.

Loading…