The properties of general Fourier partial sums of functions f ∈ C L f\in C_{L}
G. Cagareishvili, V. Tsagareishvili, and G. Tutberidze · 2025
Read The properties of general Fourier partial sums of functions f ∈ C L f\in C_{L} by G. Cagareishvili, V. Tsagareishvili, and G. Tutberidze (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-01826.txt.
Introduction
Three mathematicians, Cagareishvili, Tsagareishvili, and Tutberidze, study how well a classic series method can rebuild a smooth function from its parts. They look at sums built from general orthonormal systems, a broad family of wave-like building blocks. Their 2025 work limits the study to a named class of functions that can be differentiated. The short abstract states the aim and scope plainly, without giving detailed findings.
Catalog record
| CREATOR | G. Cagareishvili, V. Tsagareishvili, and G. Tutberidze |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 24 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-01826
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
The properties of general Fourier partial sums of functions f ∈ C L f\in C_{L} G. Cagareishvili, V. Tsagareishvili, and G. Tutberidze Abstract . In this paper, we investigated the Fourier partial sums with respect to general orthonormal systems when the function f f belongs to some differentiable class of functions 2020 Mathematics Subject Classification. 42C10, 46B07 Key words and phrases: 1. Introduction To maintain the flow of our discussion and the clarity of the proofs for our key findings, we have compiled all relevant notations, definitions, and preliminary concepts in Section 2. As S. Banach proved that good differential properties of a function do not guarantee the a.e. boundedness of the Fourier partial sum of this function with respect to general orthonormal systems (ONS). In this paper we have studied the Fourier partial sums with respect to general ONS when function f belongs to some differentiable class of functions. We also prove that this result is, in a sense, sharp. See Theorem 2 . Concerning some recent such applications equipped with the present authors, we refer to the PhD. theses [ 17 ] and [ 21 ] and the references therein. In particular, some important engineering and industrial applications are investigated in these PhD. theses, e.g. concerning structural Health Monitoring, Artificial Intelligence, Neural Networks, Signal processing, Operational Model Analysis (OMA), Damage Detection of Bridges with focus on the impressive Hålogaland Bridge in Narvik. For our investigation, it is interesting that according to Menchov’s and Banach’s results, the convergence of general orthonormal series and the convergence of general Fourier series for functions from some differential classes are different problems. In the first case, the coefficients of the orthonormal series play a decisive role. In the second case, the fact that a function f (f\neq 0) belongs to any differential class does not guarantee the convergence of its Fourier series for general ONSs. Therefore, for the Fourier series, with respect to an ONS, to be convergent, one should impose additional conditions on the functions \varphi_{n} of the ONS (\varphi_{n}). Our main problem is to find the conditions for the functions \varphi_{n} of an ONS (\varphi_{n}) for the Fourier series of every function of Lip1 classes to be bounded the Fourier partial sums. Another motivation for this paper is to use our main results to obtain some new results concerning convergence/divergence of general Fourier series. Some other results for this case can be found in [ 3 , 5 , 6 , 7 , 10 , 11 , 13 , 14 , 15 , 16 , 18 , 19 , 20 ] . See also the book [ 12 ] and monographs [ 1 , 8 , 9 ] . The main results Theorems 1 and 2 are presented and proved in Section 3. 2. Auxiliary definitions and results By Lip1 we denote the class of functions f from C(0,1) , for which \|f(x)-f(y)\|=O(h),\text{ when }\max|x-y|\leq h. Let C_{L} be the class of functions f if \frac{d}{dx}f\in Lip1. Suppose that f\in L_{2} be an arbitrary function and (\varphi_{n}) be ONS on [0,1] . Then the numbers C_{n}(f)=\int_{0}^{1}f(x)\varphi_{n}(x)\,dx,\quad n=1,2,\dots\,, are the Fourier coefficients of function f with respect to the system (\varphi_{n}) and \sum_{n=1}^{\infty}C_{n}(f)\varphi_{n}(x) (1) is the Fourier series of this function f . The general partial sum S_{n}(x,f) of the series ( 1 ) is defined as follows \displaystyle S_{n}(x,f)=\sum_{k=1}^{n}C_{k}(f)\varphi_{k}(x). (2) Next, we can define M_{n}(x) as follows \left(x\in[0,1]\right) M_{n}(x)=\frac{1}{n}\sum_{i=1}^{n-1}\left|\int_{0}^{\frac{i}{n}}Q_{n}(u,x)du\right|, (3) where Q_{n}(u,x)=\sum_{k=1}^{n}g_{k}(u)\varphi_{k}(x) (4) and \displaystyle g_{k}(u)=\int_{0}^{u}\varphi_{k}(t)dt. (5) Lemma 1 . (see [ 16 ] ) Let (\varphi_{n}) be an arbitrary ONS on [0,1] . Then \frac{1}{n^{2}}\sum_{k=1}^{n}\varphi_{k}^{2}(x)=O(1)n^{-\frac{1}{2}}\;\;\text{a.e. on $[0,1]$}. Notation. Let G=[0,1]\setminus F, where \displaystyle F=\left\{x\in[0,1]\left|\lim\limits_{n\to\infty}\frac{1}{n^{2}}\sum_{k=1}^{n}\varphi_{k}^{2}(x)=\infty\right.\right\}. It is easy to show that \left|F\right|=0\text{ and }\left|G\right|=1. Lemma 2 . (see [ 16 ] ) Suppose that B_{n}(u,x)=\sum_{k=1}^{n}\varphi_{k}(u)\varphi_{k}(x). (6) If for some x\in[0,1] \limsup_{n\to\infty}\bigg|\int_{0}^{1}B_{n}(u,x)\,du\bigg|=+\infty, (7) then for the function q(u)=1 (u\in[0,1]) \displaystyle\limsup_{n\to\infty}|S_{n}(x,q)|=+\infty, where \displaystyle S_{n}(x,q)=\sum_{k=1}^{n}C_{k}(q)\varphi_{k}(x)\text{ and }C_{k}(g)=\int_{0}^{1}q(u)\varphi_{k}(u)du=\int_{0}^{1}\varphi_{n}(u)du. Proof. Indeed, as C_{n}(q)=\int_{0}^{1}\varphi_{n}(u)du we get \displaystyle\int_{0}^{1}B_{n}(u,x)du=\sum_{k=1}^{n}\int_{0}^{1}\varphi_{k}(u)du\varphi_{k}(x)=\sum_{k=1}^{n}C_{k}(q)\varphi_{k}(x)=S_{n}(x,q). From the last equality and ( 7 ) we conclude that Lemma 2 is valid. ∎ Lemma 3 . (see [ 16 ] ) For any i (i=1,\dots,n) and x\in[0,1] \int_{\frac{i-1}{n}}^{\frac{i}{n}}|Q_{n}(u,x)|\,du\leq\frac{1}{n}\bigg(\sum_{k=1}^{n}\varphi_{k}^{2}(x)\bigg)^{\frac{1}{2}} holds (see ( 4 )). Proof. By using ( 5 ) and Bessel inequality, \displaystyle\sum_{k=1}^{\infty}g_{k}^{2}(u)=\sum_{k=1}^{\infty}\left(\int_{0}^{u}\varphi_{k}(t)dt\right)^{2}\leq\int_{0}^{u}dt\leq 1 (8) where u\in[0,1]. Using ( 8 ), the Cauchy and Hölder inequalities, we get \left(i=1,2,\dots,n\right) \displaystyle\int_{\frac{i-1}{n}}^{\frac{i}{n}}|Q_{n}(u,x)|\,du \displaystyle\leq \displaystyle\frac{1}{\sqrt{n}}\left(\int_{\frac{i-1}{n}}^{\frac{i}{n}}Q_{n}^{2}(u,x)\,du\right)^{\frac{1}{2}}=\frac{1}{\sqrt{n}}\left(\int_{\frac{i-1}{n}}^{\frac{i}{n}}\left(\sum_{k=1}^{n}g_{k}(u)\varphi_{k}(x)\right)^{2}\,du\right)^{\frac{1}{2}} \displaystyle\leq \displaystyle\frac{1}{\sqrt{n}}\left(\int_{\frac{i-1}{n}}^{\frac{i}{n}}\sum\limits_{k=1}^{n}g_{k}^{2}(u)du\sum_{k=1}^{n}\varphi_{k}^{2}(x)\right)^{\frac{1}{2}}\leq\frac{1}{n}\left(\sum_{k=1}^{n}\varphi_{k}^{2}(x)\right)^{\frac{1}{2}}. The Lemma 3 is proved. ∎ Lemma 4 . (see [ 16 ] ) Let (\varphi_{n}) be an ONS on [0,1] and f\in C_{L} , then S_{n}(x,f)=f(1)\int_{0}^{1}B_{n}(u,x)\,du-\int_{0}^{1}f^{\prime}(x)Q_{n}(u,x)\,du. (9) Proof. Integrating by parts, we obtain C_{n}(f)=\int_{0}^{1}f(u)\varphi_{k}(u)\,du=f(1)\int_{0}^{1}\varphi_{k}(u)\,du-\int_{0}^{1}f^{\prime}(u)g_{n}(u)\,du. Therefore, \displaystyle S_{n}(x,f) \displaystyle=\sum_{k=1}^{n}C_{k}(f)\varphi_{k}(x) \displaystyle=f(1)\int_{0}^{1}\sum_{k=1}^{n}\varphi_{k}(u)\varphi_{k}(x)\,du-\int_{0}^{1}f^{\prime}(u)\sum_{k=1}^{n}g_{k}(u)\varphi_{k}(x)\,du \displaystyle=f(1)\int_{0}^{1}B_{n}(u,x)\,du-\int_{0}^{1}f^{\prime}(u)Q_{n}(u,x)\,du. (10) From ( 10 ) we derive ( 9 ). The Lemma 4 is proved. ∎ Lemma 5 . Suppose that q\in E(\varphi) (q(u)=1,\ u\in[0,1]) and \displaystyle Q_{n}(u,x)=\sum_{k=1}^{n}g_{k}(u)\varphi_{k}(x). If for some x\in[0,1] \displaystyle\limsup_{n\to\infty}\left|\int_{0}^{1}Q_{n}(u,x)du\right|=+\infty, (11) then for the function p(u)=u (u\in[0,1]) \displaystyle\limsup_{n\to\infty}\left|S_{n}(x,p)\right|=+\infty. Proof. In equation ( 10 ) we suppose that f=p=u (u\in[0,1]), we receive \displaystyle S_{n}(x,h)=\int_{0}^{1}B_{n}(u,x)\,du-\int_{0}^{1}Q_{n}(u,x)\,du. (12) According to the condition of this lemma \displaystyle\left|\int_{0}^{1}B_{n}(u,x)\,du\right|=O(1). Consequently, using ( 11 ) and ( 12 ) we get \displaystyle\limsup_{n\to\infty}\left|S_{n}(x,f)\right|=+\infty. Lemma 5 is proved. ∎ Definition 1 . Let E(\varphi) be a set of any functions f with |S_{n}(x,f)|=O(1) (13) at the point x\in[0,1] . Theorem A (S. Banach [ 2 ] ) . For any f\in L_{2} (f\not\sim 0) there exists an ONS (\varphi_{n}) such that \displaystyle\limsup_{n\to\infty}|S_{n}(x,f)|=+\infty\ \ \ \text{a.e. on }[0,1]. Theorem B (see [ 4 ] ) . If f,F\in L_{2} , then \displaystyle\int_{0}^{1}f^{\prime}(x)F(x)\,dx \displaystyle=n\sum_{i=1}^{n-1}\int_{\frac{i-1}{n}}^{\frac{i}{n}}\left(f(x)-f\left(x+\frac{1}{n}\right)\right)\,dx\int_{0}^{\frac{i}{n}}F(x)\,dx \displaystyle\qquad+n\sum_{i=1}^{n-1}\int_{\frac{i-1}{n}}^{\frac{i}{n}}\int_{\frac{i-1}{n}}^{\frac{i}{n}}\left(f(x)-f(u)\right)\,duF(x)\,dx \displaystyle\qquad+n\int_{1-\frac{1}{n}}^{1}f(x)\,dx\int_{0}^{1}F(x)\,dx. (14) 3. The main problems From Theorem A it follows that even for function g(x)=1 (x\in[0,1]) there exists an ONS (\varphi_{n}) such that \limsup_{n\to\infty}|S_{n}(x,g)|=+\infty a.e. on [0,1] . Theorem 1 . Let \left(\varphi_{n}\right) be an ONS on [0,1] and p,q\in E\left(\varphi\right), where p(u)=u,q(u)=1,u\in\left[0,1\right]. If for x\in G M_{n}(x)=O(1), then for any f\in C_{L}, S_{n}(x,f)=O(1). Proof. If we substitute F(x)=Q_{n}(u,x) and f=f^{\prime} in ( 14 ), we obtain \displaystyle\int_{0}^{1}f^{\prime}(u)Q_{n}(u,x)\,du \displaystyle=n\sum_{i=1}^{n-1}\int_{\frac{i-1}{n}}^{\frac{i}{n}}\left(f^{\prime}(u)-f^{\prime}\left(u+\frac{1}{n}\right)\right)\,du\int_{0}^{\frac{i}{n}}Q_{n}(u,x)\,du \displaystyle\qquad+n\sum_{i=1}^{n-1}\int_{\frac{i-1}{n}}^{\frac{i}{n}}\int_{\frac{i-1}{n}}^{\frac{i}{n}}\left(f^{\prime}(u)-f^{\prime}(v)\right)\,dv\;Q_{n}(u,x)\,du \displaystyle\qquad+n\int_{1-\frac{1}{n}}^{1}f^{\prime}(u)\,du\;\int_{0}^{1}Q_{n}(u,x)\,du \displaystyle=I_{1}+I_{2}+I_{3}. (15) Due to ( 3 ) and the fact that f^{\prime}\in C_{L} , we get (\Delta_{in}=[\frac{i-1}{n},\frac{i}{n}]) \displaystyle|I_{1}| \displaystyle\leq O\left(\frac{1}{n}\right)n\sum_{i=1}^{n-1}\int_{\Delta_{in}}\left|\int_{0}^{\frac{i}{n}}Q_{n}\left(u,x\right)du\right|dv\leq\frac{O(1)}{n}\sum_{i=1}^{n-1}\left|\int_{0}^{\frac{i}{n}}Q_{n}\left(u,x\right)du\right| (16) \displaystyle=O\left(1\right)M_{n}(x)=O(1). Next, since f^{\prime}\in C_{L} and x\in G , by using inequality ( 8 ) from Lemma 3 and by using Hölder’s and Parseval’s \left(\sum\limits_{k=1}^{n}g_{k}^{2}(u)\leq 1\right) inequalities, we have \displaystyle|I_{2}| \displaystyle\leq O\left(\frac{1}{n^{2}}\right)n\sum_{i=1}^{n}\int_{\Delta_{in}}\left|Q_{n}(u,x)\right|du=O\left(1\right)\frac{1}{n}\int_{0}^{1}\left|Q_{n}\left(u,x\right)\right|du (17) \displaystyle=O(1)\frac{1}{n}\left(\int_{0}^{1}Q_{n}^{2}(u,x)du\right)^{\frac{1}{2}}=O(1)\frac{1}{n}\left(\int_{0}^{1}\left(\sum_{k=1}^{n}g_{k}(u)\varphi_{k}(x)\right)^{2}du\right)^{\frac{1}{2}} \displaystyle=O(1)\frac{1}{n}\left(\int_{0}^{1}\sum_{k=1}^{n}g_{k}^{2}(u)du\sum_{k=1}^{n}\varphi_{k}^{2}(x)\right)^{\frac{1}{2}}=O(1)\frac{1}{n}\left(\sum_{k=1}^{n}\varphi_{k}^{2}(x)\right)^{\frac{1}{2}} \displaystyle=O(1)\left(\frac{1}{n^{2}}\sum_{k=1}^{n}\varphi_{k}^{2}(x)\right)^{\frac{1}{2}}=O(1). Further, in ( 10 ) we suppose f(u)=p(u)=u and q(u)=1, u\in[0,1]. We receive \displaystyle S_{n}(x,p)=\int_{0}^{1}B_{n}(u,x)du-\int_{0}^{1}Q_{n}(u,x)du. From here as \displaystyle\int_{0}^{1}B_{n}(u,x)du=S_{n}(x,q). we have (see ( 6 )) \displaystyle S_{n}(x,p)=S_{n}(x,q)-\int_{0}^{1}Q_{n}(u,x)du. Consequently if p,q\in E(\varphi), we have \displaystyle|I_{3}| \displaystyle\leq \displaystyle n\int_{1-\frac{1}{n}}^{1}|f^{\prime}(u)|\,du\;\bigg|\int_{0}^{1}Q_{n}(u,x)\,du\bigg| (18) \displaystyle\leq \displaystyle n\ \frac{1}{n}\ \max_{i\in[0,1]}|f^{\prime}(u)|\bigg|\int_{0}^{1}Q_{n}(u,x)\,du\bigg|=O(1). Taking the evaluations of I_{1}, I_{2}, and I_{3} into account in ( 15 ) we conclude \bigg|\int_{0}^{1}f^{\prime}(u)Q_{n}(u,x)\,du\bigg|=O(1). (19) Finally, according to the conditions of this theorem and ( 9 ), we can deduce that S_{n}(x,f)=O(1). Theorem 1 is completely proved. ∎ Theorem 2 . Let (\varphi_{n}) be an ONS on [0,1] . If for some t\in G , \limsup_{n\to\infty}M_{n}(t)=+\infty, then there exists a function r^{\prime}\in C_{L} such that \limsup_{n\to\infty}|S_{n}(t,r)|=+\infty. Proof. Firstly, according to Lemma 2 and Lemma 5 \displaystyle\left|\int_{0}^{1}B_{n}(u,t)\,du\right|=O(1)\text{ and }\left|\int_{0}^{1}G_{n}(u,t)\,du\right|=O(1) (20) otherwise, Theorem 2 is proved. We defined the sequence of functions (f_{n}) as follows: \displaystyle f_{n}(u)=\int_{0}^{u}sign\int_{0}^{y}Q_{n}(v,t)dvdy,\ \ \ n=1,2,\dots. (21) In ( 18 ) we substitute Q_{n}\left(u,x\right)=Q_{n}\left(u,t\right) and f^{\prime}(u)=f_{n}(u), then \displaystyle\int_{0}^{1}f_{n}(u)Q_{n}(u,t)du \displaystyle= \displaystyle\sum_{i=1}^{n-1}\left(f_{n}\left(\frac{i}{n}\right)-f_{n}\left(\frac{i+1}{n}\right)\right)\int_{0}^{\frac{i}{n}}Q_{n}(u,t)du (22) \displaystyle+ \displaystyle\sum_{i=1}^{n}\int_{\frac{i-1}{n}}^{\frac{i}{n}}\left(f_{n}\left(u\right)-f_{n}\left(\frac{i}{n}\right)\right)Q_{n}(u,t)du \displaystyle+ \displaystyle f_{n}\left(1\right)\int_{0}^{1}Q_{n}(u,t)du=S_{1}+S_{2}+S_{3}. By using (