Generating function transformation

In mathematics, a transformation of a sequence's generating function provides a method of converting the generating function for one sequence into a generating function enumerating another. These transformations typically involve integral formulas applied to a sequence generating function (see integral transformations) or weighted sums over the higher-order derivatives of these functions (see derivative transformations).

Given a sequence, ${\displaystyle \{f_{n}\}_{n=0}^{\infty }}$, the ordinary generating function (OGF) of the sequence, denoted ${\displaystyle F(z)}$, and the exponential generating function (EGF) of the sequence, denoted ${\displaystyle {\widehat {F}}(z)}$, are defined by the formal power series

${\displaystyle F(z)=\sum _{n=0}^{\infty }f_{n}z^{n}=f_{0}+f_{1}z+f_{2}z^{2}+\cdots }$

${\displaystyle {\widehat {F}}(z)=\sum _{n=0}^{\infty }{\frac {f_{n}}{n!}}z^{n}={\frac {f_{0}}{0!}}+{\frac {f_{1}}{1!}}z+{\frac {f_{2}}{2!}}z^{2}+\cdots .}$

In this article, we use the convention that the ordinary (exponential) generating function for a sequence ${\displaystyle \{f_{n}\}}$ is denoted by the uppercase function ${\displaystyle F(z)}$ / ${\displaystyle {\widehat {F}}(z)}$ for some fixed or formal ${\displaystyle z}$ when the context of this notation is clear. Additionally, we use the bracket notation for coefficient extraction from the Concrete Mathematics reference which is given by ${\displaystyle [z^{n}]F(z):=f_{n}}$. The main article gives examples of generating functions for many sequences. Other examples of generating function variants include Dirichlet generating functions (DGFs), Lambert series, and Newton series. In this article we focus on transformations of generating functions in mathematics and keep a running list of useful transformations and transformation formulas.

Series multisection provides formulas for generating functions enumerating the sequence ${\displaystyle \{f_{an+b}\}}$ given an ordinary generating function ${\displaystyle F(z)}$ where ${\displaystyle a,b\in \mathbb {N} }$, ${\displaystyle a\geq 2}$, and ${\displaystyle 0\leq b<a}$. In the first two cases where ${\displaystyle (a,b):=(2,0),(2,1)}$, we can expand these arithmetic progression generating functions directly in terms of ${\displaystyle F(z)}$:

${\displaystyle \sum _{n\geq 0}f_{2n}z^{2n}={\frac {1}{2}}\left(F(z)+F(-z)\right)}$

${\displaystyle \sum _{n\geq 0}f_{2n+1}z^{2n+1}={\frac {1}{2}}\left(F(z)-F(-z)\right).}$

More generally, suppose that ${\displaystyle a\geq 3}$ and that ${\displaystyle \omega _{a}:=\exp \left({\frac {2\pi \imath }{a}}\right)}$ denotes the ${\displaystyle a^{th}}$ primitive root of unity. Then we have the following formula,^[1] often known as the root of unity filter:

${\displaystyle \sum _{n\geq 0}f_{an+b}z^{an+b}={\frac {1}{a}}\times \sum _{m=0}^{a-1}\omega _{a}^{-mb}F\left(\omega _{a}^{m}z\right).}$

For integers ${\displaystyle m\geq 1}$, another useful formula providing somewhat reversed floored arithmetic progressions are generated by the identity^[2]

${\displaystyle \sum _{n\geq 0}f_{\lfloor {\frac {n}{m}}\rfloor }z^{n}={\frac {1-z^{m}}{1-z}}F(z^{m})=\left(1+z+\cdots +z^{m-2}+z^{m-1}\right)F(z^{m}).}$

The exponential Bell polynomials, ${\displaystyle B_{n,k}(x_{1},\ldots ,x_{n}):=n!\cdot [t^{n}u^{k}]\Phi (t,u)}$, are defined by the exponential generating function^[3]

${\displaystyle \Phi (t,u)=\exp \left(u\times \sum _{m\geq 1}x_{m}{\frac {t^{m}}{m!}}\right)=1+\sum _{n\geq 1}\left\{\sum _{k=1}^{n}B_{n,k}(x_{1},x_{2},\ldots )u^{k}\right\}{\frac {t^{n}}{n!}}.}$

The next formulas for powers, logarithms, and compositions of formal power series are expanded by these polynomials with variables in the coefficients of the original generating functions.^[4][5] The formula for the exponential of a generating function is given implicitly through the Bell polynomials by the EGF for these polynomials defined in the previous formula for some sequence of ${\displaystyle \{x_{i}\}}$.

The power series for the reciprocal of a generating function, ${\displaystyle F(z)}$, is expanded by

${\displaystyle {\frac {1}{F(z)}}={\frac {1}{f_{0}}}-{\frac {f_{1}}{f_{0}^{2}}}z+{\frac {\left(f_{1}^{2}-f_{0}f_{2}\right)}{f_{0}^{3}}}z^{2}-{\frac {f_{1}^{3}-2f_{0}f_{1}f_{2}+f_{0}^{2}f_{3}}{f_{0}^{4}}}+\cdots .}$

If we let ${\displaystyle b_{n}:=[z^{n}]1/F(z)}$ denote the coefficients in the expansion of the reciprocal generating function, then we have the following recurrence relation:

${\displaystyle b_{n}=-{\frac {1}{f_{0}}}\left(f_{1}b_{n-1}+f_{2}b_{n-2}+\cdots +f_{n}b_{0}\right),n\geq 1.}$

Let ${\displaystyle m\in \mathbb {C} }$ be fixed, suppose that ${\displaystyle f_{0}=1}$, and denote ${\displaystyle b_{n}^{(m)}:=[z^{n}]F(z)^{m}}$. Then we have a series expansion for ${\displaystyle F(z)^{m}}$ given by

${\displaystyle F(z)^{m}=1+mf_{1}z+m\left((m-1)f_{1}^{2}+2f_{2}\right){\frac {z^{2}}{2}}+\left(m(m-1)(m-2)f_{1}^{3}+6m(m-1)f_{2}+6mf_{3}\right){\frac {z^{3}}{6}}+\cdots ,}$

and the coefficients ${\displaystyle b_{n}^{(m)}}$ satisfy a recurrence relation of the form

${\displaystyle n\cdot b_{n}^{(m)}=(m-n+1)f_{1}b_{n-1}^{(m)}+(2m-n+2)f_{2}b_{n-2}^{(m)}+\cdots +((n-1)m-1)f_{n-1}b_{1}^{(m)}+nmf_{n},n\geq 1.}$

Another formula for the coefficients, ${\displaystyle b_{n}^{(m)}}$, is expanded by the Bell polynomials as

${\displaystyle F(z)^{m}=f_{0}^{m}+\sum _{n\geq 1}\left(\sum _{1\leq k\leq n}(m)_{k}f_{0}^{m-k}B_{n,k}(f_{1}\cdot 1!,f_{2}\cdot 2!,\ldots )\right){\frac {z^{n}}{n!}},}$

where ${\displaystyle (r)_{n}}$ denotes the Pochhammer symbol.

If we let ${\displaystyle f_{0}=1}$ and define ${\displaystyle q_{n}:=[z^{n}]\log F(z)}$, then we have a power series expansion for the composite generating function given by

${\displaystyle \log F(z)=f_{1}+\left(2f_{2}-f_{1}^{2}\right){\frac {z}{2}}+\left(3f_{3}-3f_{1}f_{2}+f_{1}^{3}\right){\frac {z^{2}}{3}}+\cdots ,}$

where the coefficients, ${\displaystyle q_{n}}$, in the previous expansion satisfy the recurrence relation given by

${\displaystyle n\cdot q_{n}=nf_{n}-(n-1)f_{1}q_{n-1}-(n-2)f_{2}q_{n-2}-\cdots -f_{n-1}q_{1},}$

and a corresponding formula expanded by the Bell polynomials in the form of the power series coefficients of the following generating function:

${\displaystyle \log F(z)=\sum _{n\geq 1}\left(\sum _{1\leq k\leq n}(-1)^{k-1}(k-1)!B_{n,k}(f_{1}\cdot 1!,f_{2}\cdot 2!,\ldots )\right){\frac {z^{n}}{n!}}.}$

Let ${\displaystyle {\widehat {F}}(z)}$ denote the EGF of the sequence, ${\displaystyle \{f_{n}\}_{n\geq 0}}$, and suppose that ${\displaystyle {\widehat {G}}(z)}$ is the EGF of the sequence, ${\displaystyle \{g_{n}\}_{n\geq 0}}$. Faà di Bruno's formula implies that the sequence, ${\displaystyle \{h_{n}\}_{n\geq 0}}$, generated by the composition ${\displaystyle {\widehat {H}}(z):={\widehat {F}}({\widehat {G}}(z))}$, can be expressed in terms of the exponential Bell polynomials as follows:

${\displaystyle h_{n}=\sum _{1\leq k\leq n}f_{k}\cdot B_{n,k}(g_{1},g_{2},\cdots ,g_{n-k+1})+f_{0}\cdot \delta _{n,0}.}$

We have the following integral formulas for ${\displaystyle a,b\in \mathbb {Z} ^{+}}$ which can be applied termwise with respect to ${\displaystyle z}$ when ${\displaystyle z}$ is taken to be any formal power series variable:^[6]

${\displaystyle \sum _{n\geq 0}f_{n}z^{n}=\int _{0}^{\infty }{\widehat {F}}(tz)e^{-t}dt=z^{-1}{\mathcal {L}}[{\widehat {F}}](z^{-1})}$

${\displaystyle \sum _{n\geq 0}\Gamma (an+b)\cdot f_{n}z^{n}=\int _{0}^{\infty }t^{b-1}e^{-t}F(t^{a}z)dt.}$

${\displaystyle \sum _{n\geq 0}{\frac {f_{n}}{n!}}z^{n}={\frac {1}{2\pi }}\int _{-\pi }^{\pi }F\left(ze^{-\imath \vartheta }\right)e^{e^{\imath \vartheta }}d\vartheta .}$

Notice that the first and last of these integral formulas are used to convert between the EGF to the OGF of a sequence, and from the OGF to the EGF of a sequence whenever these integrals are convergent.

The first integral formula corresponds to the Laplace transform (or sometimes the formal Laplace–Borel transformation) of generating functions, denoted by ${\displaystyle {\mathcal {L}}[F](z)}$, defined in.^[7] Other integral representations for the gamma function in the second of the previous formulas can of course also be used to construct similar integral transformations. One particular formula results in the case of the double factorial function example given immediately below in this section. The last integral formula is compared to Hankel's loop integral for the reciprocal gamma function applied termwise to the power series for ${\displaystyle F(z)}$.

The single factorial function, ${\displaystyle (2n)!}$, is expressed as a product of two double factorial functions of the form

${\displaystyle (2n)!=(2n)!!\times (2n-1)!!={\frac {4^{n}\cdot n!}{\sqrt {\pi }}}\times \Gamma \left(n+{\frac {1}{2}}\right),}$

where an integral for the double factorial function, or rational gamma function, is given by

${\displaystyle {\frac {1}{2}}\cdot (2n-1)!!={\frac {2^{n}}{\sqrt {4\pi }}}\Gamma \left(n+{\frac {1}{2}}\right)={\frac {1}{\sqrt {2\pi }}}\times \int _{0}^{\infty }e^{-t^{2}/2}t^{2n}\,dt,}$

for natural numbers ${\displaystyle n\geq 0}$. This integral representation of ${\displaystyle (2n-1)!!}$ then implies that for fixed non-zero ${\displaystyle q\in \mathbb {C} }$ and any integral powers ${\displaystyle k\geq 0}$, we have the formula

${\displaystyle {\frac {\log(q)^{k}}{k!}}={\frac {1}{(2k)!}}\times \left[\int _{0}^{\infty }{\frac {2e^{-t^{2}/2}}{\sqrt {2\pi }}}\left({\sqrt {2\log(q)}}\cdot t\right)^{2k}\,dt\right].}$

Thus for any prescribed integer ${\displaystyle j\geq 0}$, we can use the previous integral representation together with the formula for extracting arithmetic progressions from a sequence OGF given above, to formulate the next integral representation for the so-termed modified Stirling number EGF as

${\displaystyle \sum _{n\geq 0}\left\{{\begin{matrix}2n\\j\end{matrix}}\right\}{\frac {\log(q)^{n}}{n!}}=\int _{0}^{\infty }{\frac {e^{-t^{2}/2}}{{\sqrt {2\pi }}\cdot j!}}\left[\sum _{b=\pm 1}\left(e^{b{\sqrt {2\log(q)}}\cdot t}-1\right)^{j}\right]dt,}$

which is convergent provided suitable conditions on the parameter ${\displaystyle 0<|q|<1}$.^[8]

For fixed non-zero ${\displaystyle c,z\in \mathbb {C} }$ defined such that ${\displaystyle |cz|<1}$, let the geometric series over the non-negative integral powers of ${\displaystyle (cz)^{n}}$ be denoted by ${\displaystyle G(z):=1/(1-cz)}$. The corresponding higher-order ${\displaystyle j^{th}}$ derivatives of the geometric series with respect to ${\displaystyle z}$ are denoted by the sequence of functions

${\displaystyle G_{j}(z):={\frac {(cz)^{j}}{1-cz}}\times \left({\frac {d}{dz}}\right)^{(j)}\left[G(z)\right],}$

for non-negative integers ${\displaystyle j\geq 0}$. These ${\displaystyle j^{th}}$ derivatives of the ordinary geometric series can be shown, for example by induction, to satisfy an explicit closed-form formula given by

${\displaystyle G_{j}(z)={\frac {(cz)^{j}\cdot j!}{(1-cz)^{j+2}}},}$

for any ${\displaystyle j\geq 0}$ whenever ${\displaystyle |cz|<1}$. As an example of the third OGF ${\displaystyle \longmapsto }$ EGF conversion formula cited above, we can compute the following corresponding exponential forms of the generating functions ${\displaystyle G_{j}(z)}$:

${\displaystyle {\widehat {G}}_{j}(z)={\frac {1}{2\pi }}\int _{-\pi }^{+\pi }G_{j}\left(ze^{-\imath t}\right)e^{e^{\imath t}}dt={\frac {(cz)^{j}e^{cz}}{(j+1)}}\left(j+1+z\right).}$

Fractional integrals and fractional derivatives (see the main article) form another generalized class of integration and differentiation operations that can be applied to the OGF of a sequence to form the corresponding OGF of a transformed sequence. For ${\displaystyle \Re (\alpha )>0}$ we define the fractional integral operator (of order ${\displaystyle \alpha }$) by the integral transformation^[9]

${\displaystyle I^{\alpha }F(z)={\frac {1}{\Gamma (\alpha )}}\int _{0}^{z}(z-t)^{\alpha -1}F(t)dt,}$

which corresponds to the (formal) power series given by

${\displaystyle I^{\alpha }F(z)=\sum _{n\geq 0}{\frac {n!}{\Gamma (n+\alpha +1)}}f_{n}z^{n+\alpha }.}$

For fixed ${\displaystyle \alpha ,\beta \in \mathbb {C} }$ defined such that ${\displaystyle \Re (\alpha ),\Re (\beta )>0}$, we have that the operators ${\displaystyle I^{\alpha }I^{\beta }=I^{\alpha +\beta }}$. Moreover, for fixed ${\displaystyle \alpha \in \mathbb {C} }$ and integers ${\displaystyle n}$ satisfying ${\displaystyle 0<\Re (\alpha )<n}$ we can define the notion of the fractional derivative satisfying the properties that

${\displaystyle D^{\alpha }F(z)={\frac {d^{(n)}}{dz^{(n)}}}I^{n-\alpha }F(z),}$

and

${\displaystyle D^{k}I^{\alpha }=D^{n}I^{\alpha +n-k}}$ for ${\displaystyle k=1,2,\ldots ,n,}$

where we have the semigroup property that ${\displaystyle D^{\alpha }D^{\beta }=D^{\alpha +\beta }}$ only when none of ${\displaystyle \alpha ,\beta ,\alpha +\beta }$ is integer-valued.

For fixed ${\displaystyle s\in \mathbb {Z} ^{+}}$, we have that (compare to the special case of the integral formula for the Nielsen generalized polylogarithm function defined in^[10]) ^[11]

${\displaystyle \sum _{n\geq 0}{\frac {f_{n}}{(n+1)^{s}}}z^{n}={\frac {(-1)^{s-1}}{(s-1)!}}\int _{0}^{1}\log ^{s-1}(t)F(tz)dt.}$

Notice that if we set ${\displaystyle g_{n}\equiv f_{n+1}}$, the integral with respect to the generating function, ${\displaystyle G(z)}$, in the last equation when ${\displaystyle z\equiv 1}$ corresponds to the Dirichlet generating function, or DGF, ${\displaystyle {\widetilde {F}}(s)}$, of the sequence of ${\displaystyle \{f_{n}\}}$ provided that the integral converges. This class of polylogarithm-related integral transformations is related to the derivative-based zeta series transformations defined in the next sections.

For fixed non-zero ${\displaystyle q,c,z\in \mathbb {C} }$ such that ${\displaystyle |q|<1}$ and ${\displaystyle |cz|<1}$, we have the following integral representations for the so-termed square series generating function associated with the sequence ${\displaystyle \{f_{n}\}}$, which can be integrated termwise with respect to ${\displaystyle z}$:^[12]

${\displaystyle \sum _{n\geq 0}q^{n^{2}}f_{n}\cdot (cz)^{n}={\frac {1}{\sqrt {2\pi }}}\int _{0}^{\infty }e^{-t^{2}/2}\left[F\left(e^{t{\sqrt {2\log(q)}}}\cdot cz\right)+F\left(e^{-t{\sqrt {2\log(q)}}}\cdot cz\right)\right]dt.}$

This result, which is proved in the reference, follows from a variant of the double factorial function transformation integral for the Stirling numbers of the second kind given as an example above. In particular, since

${\displaystyle q^{n^{2}}=\exp(n^{2}\cdot \log(q))=1+n^{2}\log(q)+n^{4}{\frac {\log(q)^{2}}{2!}}+n^{6}{\frac {\log(q)^{3}}{3!}}+\cdots ,}$

we can use a variant of the positive-order derivative-based OGF transformations defined in the next sections involving the Stirling numbers of the second kind to obtain an integral formula for the generating function of the sequence, ${\displaystyle \left\{S(2n,j)/n!\right\}}$, and then perform a sum over the ${\displaystyle j^{th}}$ derivatives of the formal OGF, ${\displaystyle F(z)}$ to obtain the result in the previous equation where the arithmetic progression generating function at hand is denoted by

${\displaystyle \sum _{n\geq 0}\left\{{\begin{matrix}2n\\j\end{matrix}}\right\}{\frac {z^{2n}}{(2n)!}}={\frac {1}{2j!}}\left((e^{z}-1)^{j}+(e^{-z}-1)^{j}\right),}$

for each fixed ${\displaystyle j\in \mathbb {N} }$.

We have an integral representation for the Hadamard product of two generating functions, ${\displaystyle F(z)}$ and ${\displaystyle G(z)}$, stated in the following form:

${\displaystyle (F\odot G)(z):=\sum _{n\geq 0}f_{n}g_{n}z^{n}={\frac {1}{2\pi }}\int _{0}^{2\pi }F\left({\sqrt {z}}e^{it}\right)G\left({\sqrt {z}}e^{-it}\right)dt,}$

where i is the imaginary unit.

More information about Hadamard products as diagonal generating functions of multivariate sequences and/or generating functions and the classes of generating functions these diagonal OGFs belong to is found in Stanley's book.^[13] The reference also provides nested coefficient extraction formulas of the form

${\displaystyle \operatorname {diag} \left(F_{1}\cdots F_{k}\right):=\sum _{n\geq 0}f_{1,n}\cdots f_{k,n}z^{n}=[x_{k-1}^{0}\cdots x_{2}^{0}x_{1}^{0}]F_{k}\left({\frac {z}{x_{k-1}}}\right)F_{k-1}\left({\frac {x_{k-1}}{x_{k-2}}}\right)\cdots F_{2}\left({\frac {x_{2}}{x_{1}}}\right)F_{1}(x_{1}),}$

which are particularly useful in the cases where the component sequence generating functions, ${\displaystyle F_{i}(z)}$, can be expanded in a Laurent series, or fractional series, in ${\displaystyle z}$, such as in the special case where all of the component generating functions are rational, which leads to an algebraic form of the corresponding diagonal generating function.

In general, the Hadamard product of two rational generating functions is itself rational.^[14] This is seen by noticing that the coefficients of a rational generating function form quasi-polynomial terms of the form

${\displaystyle f_{n}=p_{1}(n)\rho _{1}^{n}+\cdots +p_{\ell }(n)\rho _{\ell }^{n},}$

where the reciprocal roots, ${\displaystyle \rho _{i}\in \mathbb {C} }$, are fixed scalars and where ${\displaystyle p_{i}(n)}$ is a polynomial in ${\displaystyle n}$ for all ${\displaystyle 1\leq i\leq \ell }$. For example, the Hadamard product of the two generating functions

${\displaystyle F(z):={\frac {1}{1+a_{1}z+a_{2}z^{2}}}}$

and

${\displaystyle G(z):={\frac {1}{1+b_{1}z+b_{2}z^{2}}}}$

is given by the rational generating function formula^[15]

${\displaystyle (F\odot G)(z)={\frac {1-a_{2}b_{2}z^{2}}{1-a_{1}b_{1}z+\left(a_{2}b_{1}^{2}+a_{1}^{2}b_{2}-a_{2}b_{2}\right)z^{2}-a_{1}a_{2}b_{1}b_{2}z^{3}+a_{2}^{2}b_{2}^{2}z^{4}}}.}$

Ordinary generating functions for generalized factorial functions formed as special cases of the generalized rising factorial product functions, or Pochhammer k-symbol, defined by

${\displaystyle p_{n}(\alpha ,R):=R(R+\alpha )\cdots (R+(n-1)\alpha )=\alpha ^{n}\cdot \left({\frac {R}{\alpha }}\right)_{n},}$

where ${\displaystyle R}$ is fixed, ${\displaystyle \alpha \neq 0}$, and ${\displaystyle (x)_{n}}$ denotes the Pochhammer symbol are generated (at least formally) by the Jacobi-type J-fractions (or special forms of continued fractions) established in the reference.^[16] If we let ${\displaystyle \operatorname {Conv} _{h}(\alpha ,R;z):=\operatorname {FP} _{h}(\alpha ,R;z)/\operatorname {FQ} _{h}(\alpha ,R;z)}$ denote the ${\displaystyle h^{\text{th}}}$ convergent to these infinite continued fractions where the component convergent functions are defined for all integers ${\displaystyle h\geq 2}$ by

${\displaystyle \operatorname {FP} _{h}(\alpha ,R;z)=\sum _{n=0}^{h-1}\left[\sum _{k=0}^{n}{\binom {h}{k}}\left(1-h-{\frac {R}{\alpha }}\right)_{k}\left({\frac {R}{\alpha }}\right)_{n-k}\right](\alpha z)^{n},}$

and

${\displaystyle {\begin{aligned}\operatorname {FQ} _{h}(\alpha ,R;z)&=(-\alpha z)^{h}\cdot h!\times L_{h}^{\left(R/\alpha -1\right)}\left((\alpha z)^{-1}\right)\\&=\sum _{k=0}^{h}{\binom {h}{k}}\left[\prod _{j=0}^{k-1}(R+(j-1-j)\alpha )\right](-z)^{k},\end{aligned}}}$

where ${\displaystyle L_{n}^{(\beta )}(x)}$ denotes an associated Laguerre polynomial, then we have that the ${\displaystyle h^{th}}$ convergent function, ${\displaystyle \operatorname {Conv} _{h}(\alpha ,R;z)}$, exactly enumerates the product sequences, ${\displaystyle p_{n}(\alpha ,R)}$, for all ${\displaystyle 0\leq n<2h}$. For each ${\displaystyle h\geq 2}$, the ${\displaystyle h^{th}}$ convergent function is expanded as a finite sum involving only paired reciprocals of the Laguerre polynomials in the form of

${\displaystyle \operatorname {Conv} _{h}(\alpha ,R;z)=\sum _{i=0}^{h-1}{\binom {{\frac {R}{\alpha }}+i-1}{i}}\times {\frac {(-\alpha z)^{-1}}{(i+1)\cdot L_{i}^{\left(R/\alpha -1\right)}\left((\alpha z)^{-1}\right)L_{i+1}^{\left(R/\alpha -1\right)}\left((\alpha z)^{-1}\right)}}}$

Moreover, since the single factorial function is given by both ${\displaystyle n!=p_{n}(1,1)}$ and ${\displaystyle n!=p_{n}(-1,n)}$, we can generate the single factorial function terms using the approximate rational convergent generating functions up to order ${\displaystyle 2h}$. This observation suggests an approach to approximating the exact (formal) Laplace–Borel transform usually given in terms of the integral representation from the previous section by a Hadamard product, or diagonal-coefficient, generating function. In particular, given any OGF ${\displaystyle G(z)}$ we can form the approximate Laplace transform, which is ${\displaystyle 2h}$-order accurate, by the diagonal coefficient extraction formula stated above given by

${\displaystyle {\begin{aligned}{\widetilde {\mathcal {L}}}_{h}[G](z)&:=[x^{0}]\operatorname {Conv} _{h}\left(1,1;{\frac {z}{x}}\right)G(x)\\&\ ={\frac {1}{2\pi }}\int _{0}^{2\pi }\operatorname {Conv} _{h}\left(1,1;{\sqrt {z}}e^{It}\right)G\left({\sqrt {z}}e^{-It}\right)dt.\end{aligned}}}$

Examples of sequences enumerated through these diagonal coefficient generating functions arising from the sequence factorial function multiplier provided by the rational convergent functions include

${\displaystyle {\begin{aligned}n!^{2}&=[z^{n}][x^{0}]\operatorname {Conv} _{h}\left(-1,n;{\frac {z}{x}}\right)\operatorname {Conv} _{h}\left(-1,n;x\right),h\geq n\\{\binom {2n}{n}}&=[x_{1}^{0}x_{2}^{0}z^{n}]\operatorname {Conv} _{h}\left(-2,2n;{\frac {z}{x_{2}}}\right)\operatorname {Conv} _{h}\left(-2,2n-1;{\frac {x_{2}}{x_{1}}}\right)I_{0}(2{\sqrt {x_{1}}})\\{\binom {3n}{n}}{\binom {2n}{n}}&=[x_{1}^{0}x_{2}^{0}z^{n}]\operatorname {Conv} _{h}\left(-3,3n-1;{\frac {3z}{x_{2}}}\right)\operatorname {Conv} _{h}\left(-3,3n-2;{\frac {x_{2}}{x_{1}}}\right)I_{0}(2{\sqrt {x_{1}}})\\!n&=n!\times \sum _{i=0}^{n}{\frac {(-1)^{i}}{i!}}=[z^{n}x^{0}]\left({\frac {e^{-x}}{(1-x)}}\operatorname {Conv} _{n}\left(-1,n;{\frac {z}{x}}\right)\right)\\\operatorname {af} (n)&=\sum _{k=1}^{n}(-1)^{n-k}k!=[z^{n}]\left({\frac {\operatorname {Conv} _{n}(1,1;z)-1}{1+z}}\right)\\(t-1)^{n}P_{n}\left({\frac {t+1}{t-1}}\right)&=\sum _{k=0}^{n}{\binom {n}{k}}^{2}t^{k}\\&=[x_{1}^{0}x_{2}^{0}][z^{n}]\left(\operatorname {Conv} _{n}\left(1,1;{\frac {z}{x_{1}}}\right)\operatorname {Conv} _{n}\left(1,1;{\frac {x_{1}}{x_{2}}}\right)I_{0}(2{\sqrt {t\cdot x_{2}}})I_{0}(2{\sqrt {x_{2}}})\right),n\geq 1\\(2n-1)!!&=\sum _{k=1}^{n}{\frac {(n-1)!}{(k-1)!}}k\cdot (2k-3)!!\\&=[x_{1}^{0}x_{2}^{0}x_{3}^{n-1}]\left(\operatorname {Conv} _{n}\left(1,1;{\frac {x_{3}}{x_{2}}}\right)\operatorname {Conv} _{n}\left(2,1;{\frac {x_{2}}{x_{1}}}\right){\frac {(x_{1}+1)e^{x_{1}}}{(1-x_{2})}}\right),\end{aligned}}}$

where ${\displaystyle I_{0}(z)}$ denotes a modified Bessel function, ${\displaystyle !n}$ denotes the subfactorial function, ${\displaystyle \operatorname {af} (n)}$ denotes the alternating factorial function, and ${\displaystyle P_{n}(x)}$ is a Legendre polynomial. Other examples of sequences enumerated through applications of these rational Hadamard product generating functions given in the article include the Barnes G-function, combinatorial sums involving the double factorial function, sums of powers sequences, and sequences of binomials.

For fixed ${\displaystyle k\in \mathbb {Z} ^{+}}$, we have that if the sequence OGF ${\displaystyle F(z)}$ has ${\displaystyle j^{th}}$ derivatives of all required orders for ${\displaystyle 1\leq j\leq k}$, that the positive-order zeta series transformation is given by^[17]

${\displaystyle \sum _{n\geq 0}n^{k}f_{n}z^{n}=\sum _{j=0}^{k}\left\{{\begin{matrix}k\\j\end{matrix}}\right\}z^{j}F^{(j)}(z),}$

where ${\displaystyle \scriptstyle {\left\{{\begin{matrix}n\\k\end{matrix}}\right\}}}$ denotes a Stirling number of the second kind. In particular, we have the following special case identity when ${\displaystyle f_{n}\equiv 1\forall n}$ when ${\displaystyle \scriptstyle {\left\langle {\begin{matrix}n\\m\end{matrix}}\right\rangle }}$ denotes the triangle of first-order Eulerian numbers:^[18]

${\displaystyle \sum _{n\geq 0}n^{k}z^{n}=\sum _{j=0}^{k}\left\{{\begin{matrix}k\\j\end{matrix}}\right\}{\frac {z^{j}\cdot j!}{(1-z)^{j+1}}}={\frac {1}{(1-z)^{k+1}}}\times \sum _{0\leq m<k}\left\langle {\begin{matrix}k\\m\end{matrix}}\right\rangle z^{m+1}.}$

We can also expand the negative-order zeta series transformations by a similar procedure to the above expansions given in terms of the ${\displaystyle j^{th}}$-order derivatives of some ${\displaystyle F(z)\in C^{\infty }}$ and an infinite, non-triangular set of generalized Stirling numbers in reverse, or generalized Stirling numbers of the second kind defined within this context.

In particular, for integers ${\displaystyle k,j\geq 0}$, define these generalized classes of Stirling numbers of the second kind by the formula

${\displaystyle \left\{{\begin{matrix}k+2\\j\end{matrix}}\right\}_{\ast }:={\frac {1}{j!}}\times \sum _{m=1}^{j}{\binom {j}{m}}{\frac {(-1)^{j-m}}{m^{k}}}.}$

Then for ${\displaystyle k\in \mathbb {Z} ^{+}}$ and some prescribed OGF, ${\displaystyle F(z)\in C^{\infty }}$, i.e., so that the higher-order ${\displaystyle j^{th}}$ derivatives of ${\displaystyle F(z)}$ exist for all ${\displaystyle j\geq 0}$, we have that

${\displaystyle \sum _{n\geq 1}{\frac {f_{n}}{n^{k}}}z^{n}=\sum _{j\geq 1}\left\{{\begin{matrix}k+2\\j\end{matrix}}\right\}_{\ast }z^{j}F^{(j)}(z).}$

A table of the first few zeta series transformation coefficients, ${\displaystyle \scriptstyle {\left\{{\begin{matrix}k\\j\end{matrix}}\right\}_{\ast }}}$, appears below. These weighted-harmonic-number expansions are almost identical to the known formulas for the Stirling numbers of the first kind up to the leading sign on the weighted harmonic number terms in the expansions.

k	${\displaystyle \left\{{\begin{matrix}k\\j\end{matrix}}\right\}_{\ast }\times (-1)^{j-1}j!}$
2	${\displaystyle 1}$
3	${\displaystyle H_{j}}$
4	${\displaystyle {\frac {1}{2}}\left(H_{j}^{2}+H_{j}^{(2)}\right)}$
5	${\displaystyle {\frac {1}{6}}\left(H_{j}^{3}+3H_{j}H_{j}^{(2)}+2H_{j}^{(3)}\right)}$
6	${\displaystyle {\frac {1}{24}}\left(H_{j}^{4}+6H_{j}^{2}H_{j}^{(2)}+3\left(H_{j}^{(2)}\right)^{2}+8H_{j}H_{j}^{(3)}+6H_{j}^{(4)}\right)}$

The next series related to the polylogarithm functions (the dilogarithm and trilogarithm functions, respectively), the alternating zeta function and the Riemann zeta function are formulated from the previous negative-order series results found in the references. In particular, when ${\displaystyle s:=2}$ (or equivalently, when ${\displaystyle k:=4}$ in the table above), we have the following special case series for the dilogarithm and corresponding constant value of the alternating zeta function:

${\displaystyle {\begin{aligned}{\text{Li}}_{2}(z)&=\sum _{j\geq 1}{\frac {(-1)^{j-1}}{2}}\left(H_{j}^{2}+H_{j}^{(2)}\right){\frac {z^{j}}{(1-z)^{j+1}}}\\\zeta ^{\ast }(2)&={\frac {\pi ^{2}}{12}}=\sum _{j\geq 1}{\frac {\left(H_{j}^{2}+H_{j}^{(2)}\right)}{4\cdot 2^{j}}}.\end{aligned}}}$

When ${\displaystyle s:=3}$ (or when ${\displaystyle k:=5}$ in the notation used in the previous subsection), we similarly obtain special case series for these functions given by

${\displaystyle {\begin{aligned}{\text{Li}}_{3}(z)&=\sum _{j\geq 1}{\frac {(-1)^{j-1}}{6}}\left(H_{j}^{3}+3H_{j}H_{j}^{(2)}+2H_{j}^{(3)}\right){\frac {z^{j}}{(1-z)^{j+1}}}\\\zeta ^{\ast }(3)&={\frac {3}{4}}\zeta (3)=\sum _{j\geq 1}{\frac {\left(H_{j}^{3}+3H_{j}H_{j}^{(2)}+2H_{j}^{(3)}\right)}{12\cdot 2^{j}}}\\&={\frac {1}{6}}\log(2)^{3}+\sum _{j\geq 0}{\frac {H_{j}H_{j}^{(2)}}{2^{j+1}}}.\end{aligned}}}$

It is known that the first-order harmonic numbers have a closed-form exponential generating function expanded in terms of the natural logarithm, the incomplete gamma function, and the exponential integral given by

${\displaystyle \sum _{n\geq 0}{\frac {H_{n}}{n!}}z^{n}=e^{z}\left({\mbox{E}}_{1}(z)+\gamma +\log z\right)=e^{z}\left(\Gamma (0,z)+\gamma +\log z\right).}$

Additional series representations for the r-order harmonic number exponential generating functions for integers ${\displaystyle r\geq 2}$ are formed as special cases of these negative-order derivative-based series transformation results. For example, the second-order harmonic numbers have a corresponding exponential generating function expanded by the series

${\displaystyle \sum _{n\geq 0}{\frac {H_{n}^{(2)}}{n!}}z^{n}=\sum _{j\geq 1}{\frac {H_{j}^{2}+H_{j}^{(2)}}{2\cdot (j+1)!}}z^{j}e^{z}\left(j+1+z\right).}$

A further generalization of the negative-order series transformations defined above is related to more Hurwitz-zeta-like, or Lerch-transcendent-like, generating functions. Specifically, if we define the even more general parametrized Stirling numbers of the second kind by

${\displaystyle \left\{{\begin{matrix}k+2\\j\end{matrix}}\right\}_{(\alpha ,\beta )^{\ast }}:={\frac {1}{j!}}\times \sum _{0\leq m\leq j}{\binom {j}{m}}{\frac {(-1)^{j-m}}{(\alpha m+\beta )^{k}}}}$,

for non-zero ${\displaystyle \alpha ,\beta \in \mathbb {C} }$ such that ${\displaystyle -{\frac {\beta }{\alpha }}\notin \mathbb {Z} ^{+}}$, and some fixed ${\displaystyle k\geq 1}$, we have that

${\displaystyle \sum _{n\geq 1}{\frac {f_{n}}{(\alpha n+\beta )^{k}}}z^{n}=\sum _{j\geq 1}\left\{{\begin{matrix}k+2\\j\end{matrix}}\right\}_{(\alpha ,\beta )^{\ast }}z^{j}F^{(j)}(z).}$

Moreover, for any integers ${\displaystyle u,u_{0}\geq 0}$, we have the partial series approximations to the full infinite series in the previous equation given by

${\displaystyle \sum _{n=1}^{u}{\frac {f_{n}}{(\alpha n+\beta )^{k}}}z^{n}=[w^{u}]\left(\sum _{j=1}^{u+u_{0}}\left\{{\begin{matrix}k+2\\j\end{matrix}}\right\}_{(\alpha ,\beta )^{\ast }}{\frac {(wz)^{j}F^{(j)}(wz)}{1-w}}\right).}$