Generalized typical dimension of a graded module (Q2673107)
From MaRDI portal
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | Generalized typical dimension of a graded module |
scientific article |
Statements
Generalized typical dimension of a graded module (English)
0 references
9 June 2022
0 references
Let \(\mathcal{F}\) be a field and \(T = T(X)\) the free commutative semigroup with unity generated by a finite set \(X=\{x_{1},\dots, x_{m}\}\), where the order of an element (``monomial'') \(\theta=x_{1}^{e_{1}}\dots x_{m}^{e_{m}}\in T\) (denoted by \(\mathrm{ord}\theta\)) is defined as \(\sum_{i=1}^{m}e_{i}\). For any \(s\in\mathbb{N}_{0}\), let \(T_{s} = \{\theta\in T|\mathrm{ord}\theta = s\}\). We suppose that \(T\) is linearly ordered by an ordering \(<\) such that \(1\leq\theta\) for any \(\theta\in T\) and \(\theta\theta_{1} < \theta\theta_{2}\) if \(\theta_{1} < \theta_{2}\) (\(\theta_{1}, \theta_{2}\in T\)). Let \(D\) be the vector \(\mathcal{F}\)-space with the basis \(T\) where a function of ``taking the leader'' is defined by assigning to each element \(\sum_{\theta\in T}a_{\theta}\theta\) (\(a_{\theta}\in\mathcal{F}\), \(a_{\theta}\neq 0\) for just finitely many \(\theta\)) its ``leader,'' that is, the greatest monomial \(\theta\) with \(a_{\theta}\neq 0\) (it is denoted by \(\mathbf{u}_{g}\)). Suppose that \(D\) is an \(\mathcal{F}\)-algebra where \(\mathbf{u}_{AB} = \mathbf{u}_{A}\mathbf{u}_{B}\) (\(A, B\in D\)) and \(1\theta_{1}\cdot 1\theta_{2} = 1\theta_{1}\theta_{2}\in D\) for any \(\theta_{1}, \theta_{2}\in T\). Then \(D\) is called the \textit{ring of generalized polynomials} in the indeterminates \(X=\{x_{1},\dots, x_{m}\}\) over \(\mathcal{F}\). In particular, rings of polynomials over fields and rings of differential operators over differential fields can be treated as rings of generalized polynomials. The paper under review considers finitely generated graded \(D\)-modules where \(D\) is treated as a graded ring \(D=\bigoplus_{s\in\mathbb{N}_{0}}D_{s}\) with \(D_{s}\) being the set of all homogeneous generalized polynomials of total degree \(s\) in \(x_{1},\dots, x_{m}\). The \textit{characteristic function} of a graded \(D\)-module \(M=\bigoplus_{s\in\mathbb{N}_{0}}M_{s}\) is defined as a function \(\phi_{M}^{\mathrm{gr}}:\mathbb{N}_{0}\rightarrow\mathbb{N}_{0}\) such that \(\phi_{M}^{\mathrm{gr}}(s)=\dim_{\mathcal{F}}M_{s}\). It is known (see Theorem 4.3.20 of [\textit{M. V. Kondratieva} et al., Differential and difference dimension polynomials. Dordrecht: Kluwer Academic Publishers (1999; Zbl 0930.12005)]) that for all sufficiently large \(s\in\mathbb{N}_{0}\), the function \(\phi_{M}^{\mathrm{gr}}(s)\) is polynomial; the corresponding polynomial \(\omega_{M}(s)\) has degree \(d\leq m\). Let \(F\) be a free \(D\)-module with a basis \(\{f_{1},\dots, f_{n}\}\). The set \(T_{F}=\{\theta f_{i}|\theta\in T, 1\leq i\leq n\}\) is considered with a well-ordering (``ranking'') \(<\) such that \(u\leq\theta u\) and \(\theta u\leq\theta v\) if \(u\leq v\) (\(u, v\in T_{F}\), \(\theta\in T\)). Furthermore, \(F\) is treated as a graded \(D\)-module with grading \(F_{s} = \sum_{i=1}^{n}T_{s}f_{i}\), and if \(H\) is a \(D\)-submodule of \(F\) generated by a set \(\Sigma\), then the polynomial \(\omega_{F/H}(s)\) of the graded module \(F/H\) (associated with the grading of \(F\)) is called the characteristic polynomial of \(\Sigma\). If \(d=\deg\omega_{F/H}\), then the coefficient of \(s^{d}\) in \(\omega_{F/H}\) and the number \(m-1-d\), are called the \textit{generalized typical dimension} (denoted by \(\tau_{d}(\Sigma)\)) and \textit{generalized typical codimension} of the system of \(D\)-equations \(\Sigma\), respectively. The main results of the paper are as follows. I. Let \(\Sigma\) be a system of homogeneous \(D\)-equations and \(\mathrm{ord}_{f_{j}}h\leq e_{j}\) for all \(h\in\Sigma\) (\(e_{j}\in\mathbb{N}_{0}, 1\leq j\leq n\)). Then: 1) if the codimension of \(\Sigma\) is \(0\), then \(\tau_{d}(\Sigma)\leq n\); 2) if the codimension of \(\Sigma\) is \(1\), then \(\tau_{d}(\Sigma)\leq e_{1}+\dots +e_{n}\); 3) if the codimension of \(\Sigma\) is \(2\), then \(\tau_{d}(\Sigma)\leq (e_{1}+\dots +e_{n})\max_{1\leq i\leq n} e_{i} + \prod_{i<j}e_{i}e_{j}\leq (e_{1}+\dots +e_{n})^{2}\). II. Let \(n=1\) and \(\mathrm{ord} h\leq e\) for any \(h\in\Sigma\). 1) if the codimension of \(\Sigma\) is \(3\), then \(\tau_{d}(\Sigma)\leq e^{2}(e+1)^{2}/2\); 2) if the codimension of \(\Sigma\) is \(4\), then \(\tau_{d}(\Sigma)\leq e^{2}(e+1)^{2}(3e^{4}+6e^{3}+11e^{2}+8e+8)/24\). The author also obtained a boundary for \(\tau_{d}(\Sigma)\) if the codimension of \(\Sigma\) is \(5\) and proved that in any codimension \(\tau > 0\), \(\tau_{d}(\Sigma)\leq O(e^{2^{\tau-1}})\).
0 references
0 references
0 references