For bounded vertex degrees, total degree , and a forbidden graph of bounded maximum degree, put and . Count prescribed loop pairs, double-edge pairs, and forbidden-edge pairs in the configuration model. Their means are respectively , , and . A fixed collection of disjoint prescribed pairs occurs with probability . Bounded degrees make intersecting bad configurations contribute to each fixed factorial moment. Thus the total bad-pattern count has factorial moments . The Bonferroni inequalities and uniformly bounded means give the displayed zero-pattern probability, without requiring the parameters to converge.
Coefficient extraction recovers every probability mass:
Indeed, termwise differentiation of a power series is valid inside ; evaluating the th derivative at zero kills every term except . Thus two probability generating functions with the same coefficients describe the same probability distribution, without any finite-moment assumption.
At one, approach along the real line from below. For each positive integer , termwise differentiation of a power series followed by the monotone convergence theorem gives the factorial moment
For integer , set . The limit may be infinite; it is a finite boundary value only when the corresponding factorial moment is finite. In particular, a finite first moment gives
and a finite second moment gives
The variance formula follows from . These are boundary limits of derivatives; they do not assert that the probability generating function extends smoothly through one.
For a nonnegative integer-valued random variable, its moment-generating function satisfies wherever finite. The guaranteed probability generating function domain therefore supplies finite values for real ; it does not guarantee positive exponential moments or a finite moment-generating function on a neighborhood of zero. Bounded nonnegative integer-valued random variables have polynomial probability generating functions, but their convenient behavior is not a general existence theorem.
For fixed real , the number of isolated vertices in a binomial random graph with converges to a Poisson distribution of mean . Its fixed factorial moments tend to . A union bound for small connected components, together with exclusion of intermediate components, shows that isolated vertices are asymptotically the only obstruction to connectivity. Thus the displayed limit is the zero-mass of that Poisson distribution.
Dobiński's formula 2026-10-25
For with Poisson distribution of mean one, every factorial moment satisfies . The polynomial identity
expresses powers in terms of falling factorials, where are Stirling numbers of the second kind. Substituting and using linearity of expectation proves
On the other hand, the probability mass function gives . This proves Dobiński's formula and explains why a count of set partitions is a moment of a Poisson distribution. For example, . At , take also for , obtaining . For each fixed , the ratio test proves convergence of the displayed infinite sum.
Factorial moment Created 2026-09-24 Updated 2026-11-05
The th falling factorial moment is , where .
If nonnegative integer-valued random variables have all fixed factorial moments converging to , then their convergence in distribution is to the Poisson distribution with mean . The first-moment bound gives uniform tightness, and bounds on higher moments give uniform integrability of each fixed lower moment. Every subsequential limit therefore has the indicated factorial moments. For such a limit and , the nonnegative binomial expansion and the monotone convergence theorem give . This identifies its moment-generating function as that of the Poisson distribution, uniquely determining the limit.
Fixed-number two-mode condensate Created 2026-10-20 Updated 2026-11-05
All bosons occupy the same normalized one-particle superposition with . Its occupation distribution in the original two modes is binomial, not two independent Poisson distributions. In particular, and . These factorial moments give finite-particle corrections to interaction energies. The state is also the two-mode realization of a spin coherent state with spin .
The parameter relations give and
For each fixed integer , take an -vertex set and let count triangles meeting . The event that every vertex of is counted by is exactly .
The number of such triangles is
Thus their total expected count is .
Distinct triangles have dependent containment events only when they share an edge. If that common edge meets , there are choices for it and choices for the two other vertices. If the common edge avoids , both other vertices must lie in , giving only pairs. Each pair needs five edges. Consequently its ordered Janson dependency sum is
Janson inequality gives the upper bound . For the lower bound, the avoidance of each triangle is a decreasing event. Repeated application of Harris' inequality gives
The error also tends to zero. The bounds match multiplicatively:
Summing over ordered distinct vertices yields the factorial moments
Here is an explicit justification that these moments determine the limit. For every integer , the finite alternating sums
are upper bounds for when is even and lower bounds when is odd. This follows by applying the Bonferroni inequalities to ; the full sum is precisely the indicator of . First let for a fixed truncation and use the factorial-moment limits. Then let the even and odd truncations grow. Both bounds tend to
Hence every point probability has the Poisson distribution limit, proving the Poisson threshold for vertices lying in no triangle:
This is also the factorial-moment criterion for Poisson convergence, here justified rather than left as an unproved implication.
Put and . Use the configuration model: give vertex exactly labelled half-edges and choose a uniform perfect matching of all half-edges. There are
pairings, with denoting a falling factorial. Each allowed simple graph has exactly representations: at every vertex, assign its incident graph edges bijectively to its labelled half-edges. Therefore its enumeration reduces to the probability that a pairing is simple and avoids .
Count three kinds of prescribed bad patterns. A loop pattern consists of a pair of half-edges at one vertex. A double-edge pattern consists of two half-edges at each of two distinct vertices and one of their two cross-pairings. A forbidden-edge pattern consists of one half-edge at each endpoint of an edge of . Let their respective counts in the random pairing be . Simplicity and avoidance are equivalent to , including exclusion of triple edges because those contain double-edge patterns.
A prescribed collection of disjoint half-edge pairs occurs with probability
Consequently the bad-pattern expected values are
Here and , as in the PDF. Since and has bounded maximum degree, and all three means are uniformly bounded. For the double-edge estimate, , so its contribution after division by is .
We now give the moment-counting step rather than assume independence. For each fixed , expand the factorial moment as the sum over ordered distinct bad patterns. Collections on disjoint vertex sets have, by the displayed pairing probability, joint probabilities equal to the product probabilities times . Their total contribution is
Deleting collections with a shared vertex changes the independent-product sum by , since all ordinary and forbidden vertex degrees are bounded.
For actual joint probabilities of overlapping patterns, incompatible uses of a half-edge give zero. Compatible collections with no shared paired edge lose at least one free vertex choice, without reducing the number of required pairs, so contribute . A shared paired edge can occur only between two double-edge patterns or between a double-edge pattern and a forbidden-edge pattern. In the first case their union requires at least three parallel paired edges: there are endpoint choices and probability . In the second case the endpoint pair must be an edge of : there are only choices and at least two required pairs, giving . Adding further patterns in a fixed collection does not remove this deficit: a disjoint addition contributes bounded order, while an addition to an existing cluster has only bounded half-edge choices unless it supplies new free vertices, compensated by its new required pairs. Summing the finitely many overlap types for fixed therefore gives . Thus, with ,
The Bonferroni inequalities bracket by the odd and even partial sums of . The means are bounded, so the corresponding exponential-series remainders vanish uniformly as the truncation increases. Taking and then the truncation to infinity proves the bounded-degree pairing avoidance estimate . This does not require or to converge. The exponential is bounded away from zero, so the additive error is also a relative error.
Multiplying by the pairing count and dividing by the representation factor gives
The factor multiplies the exponential; it is not part of its exponent. The TeX aid misgroups this factor, whereas the original PDF is unambiguous. The bounded-degree proof also covers the stated additional assumption .
Write , and put , for . Counting the -subsets of the events that occur gives . Set for . The Jordan-Bonferroni exact-occurrence inequalities are
Here is an integer. To prove both bounds, evaluate the truncated expression before taking its expectation. At an integer value it equals
For this is one. For , Pascal's identity, or an induction on , gives
For the original expression is zero. Thus its difference from is everywhere nonnegative for even and nonpositive for odd . Taking expectations proves the inequalities, including .
For a general nonnegative integer-valued random variable, define its factorial moment by , where is the falling factorial and . The same pointwise calculation applies to , so put
Every required moment is finite under the given limit assumption: eventual finiteness of higher factorial moments implies finiteness of the lower ones. The neighbouring upper and lower bounds differ in absolute value by
The last ratio tends to one, and the first factor involving the moment tends to zero. Since lies between every two such neighbouring bounds, each tends to it. Consequently the factorial-moment inversion is
This proves convergence of the signed series without an unjustified interchange with an expectation.
For any prescribed , take for , with . Since and , the two moment sums have tails comparable respectively to and . Thus but , as required.
Take and , as required for the displayed sparse scaling and the later normalization. A strictly balanced graph satisfies for every proper nonempty subgraph . This forces to be connected and to have no isolated vertices: otherwise a component of at least the average density, or deletion of the isolated vertices, violates strict balance. Copies here are subgraphs, not induced subgraphs.
For fixed , the factorial moment counts ordered -tuples of distinct copies all present in the binomial random graph. The contribution of vertex-disjoint copies is exactly
The factor removes the multiple embeddings of each copy coming from its graph automorphisms. For an overlapping tuple, let be its union. There are placements of each fixed union type, and the probability its edges are present is . When adding a copy meeting the earlier union in vertices and edges, the exponent changes by
It is strictly negative if this intersection is a proper nonempty subgraph, and zero only for an empty intersection or a whole copy. In particular the first vertex-overlapping copy has a proper nonempty intersection: the preceding copies are vertex-disjoint, and a connected copy contained in their union would have to equal one of them, contrary to distinctness. Its contribution is therefore strictly negative; later additions cannot undo that loss. Only finitely many union types occur for fixed , so their total contribution is . We have proved all the factorial moment limits, and the allowed factorial-moment criterion for Poisson convergence gives
This is the Poisson limit for strictly balanced subgraph counts.
The pendant-edge graph has density , whereas its proper core has density . Thus is not strictly balanced. The complete graph is strictly balanced, with 24 graph automorphisms, so at its copy count tends to .
Every copy has a unique four-vertex clique core and one selected edge joining it to a fifth vertex. For any four-set , let count the ambient edges between and its complement. Then , with mean . It is independent of the internal edges determining whether is a clique. Choose . A Chernoff bound and a union bound give
Therefore, simultaneously for every actual core, its number of pendant extensions is . Extra ambient edges cause no overcount: each selected seven-edge subgraph has exactly one core and one pendant edge. Summing over the cores yields
because is uniformly tight. This is the pendant extensions of sparse clique copies mechanism. For , neither endpoint of is an integer, and its only nonnegative integer is . Convergence in distribution consequently gives
The positive- assumption matters: the quotient printed in the question is undefined when .
Let be the number of isolated vertices and set . For fixed , the falling factorial counts ordered -tuples of distinct isolated vertices. A specified tuple is isolated exactly when all edges incident to its vertices are absent. There are such edges, so
Since and , taking logarithms gives
Thus every fixed factorial moment tends to . We use the standard factorial-moment criterion for Poisson convergence: nonnegative integer-valued random variables whose factorial moments converge, for every fixed order, to converge in distribution to . Consequently has that limiting distribution, and its mass at each fixed integer converges. In particular,
This establishes the limiting probability of at least one isolated vertex, rather than merely a bound based on its expectation.
The probability generating function encodes the population distribution in one analytic function. Its coefficients and derivative recover
provided the mean is finite. Higher derivatives at one give factorial moments, and normalization says .
Interpret the supplied birth and death expressions as transition rates: over a small time their probabilities are and . Incoming probability from and , minus both outgoing rates, gives
with . Multiplying by and summing yields
At stationarity, , so integration and normalization give
Alternatively differentiating the evolution equation at one gives , so . If , the displayed power expression is understood in its limit: , a Poisson distribution with mean .
For the probability generating function , differentiation gives
The second derivative is a factorial moment, not the raw second moment.
In the Galton-Watson process, condition on . The next generation is a sum of independent offspring counts, and so its conditional generating function is . Averaging over gives
The conditional mean is and the conditional variance is . Thus iterated expectation gives from one ancestor, and the law of total variance gives , . Iterating yields the branching-process generation moments
For the critical case , the correct limiting value is . If , nonnegative offspring are zero almost surely, so every later generation has zero mean and variance; handle this directly rather than use an ambiguous zero power. At generation zero the mean is one and variance zero.
The branching-process extinction criterion states that extinction probability is the smallest fixed point of in : probabilities of extinction by successive generations are the iterates of from zero, increasing to that fixed point. Here the equation is , factoring as . The two admissible roots are and one, so
Multiply the birth-death master equation by and sum. Equivalently, a birth changes by and a death by , so
The equation is not closed in the mean because the quadratic death rate introduces the second factorial moment.
Under the extra hypothesis that exists for every real , its power series has infinite radius of convergence. Termwise differentiation of a power series is therefore valid to every finite order. For ,
At this is the expected value of the falling factorial of :
In particular . The terms with vanish because their falling factorials contain a zero factor. These factorial moments are finite under the stated hypothesis.
For the specified Poisson distribution, the exponential series gives the probability generating function
Every derivative of is again , so the factorial moments and variance are
Also .
Substitute into the polynomial identity and apply linearity of expectation to its finite sum. The factorial moments of the Poisson distribution in part (a)(iv) are all one, giving
Thus its th moment counts set partitions: it equals the th Bell number.
Binomial factorial moments give , its mode-zero counterpart , and . Consequently the exact projected expectation is
Since and ,
The finite- correction is simply the replacement in the interaction term. Number correlations must be retained to obtain that correction.
For the binomial distribution, the binomial theorem gives
Differentiation of the probability generating function recovers the first expected value and second factorial moment:
Since , the variance is
Thus
The formulas also hold for and the degenerate choices , with derivatives read directly from the resulting polynomial.