REVIEW 4 major objections 5 minor 10 references
Incomplete (even and odd) trigonometric splines in the problems of constructing approximate solutions of second order linear differential equations
T0 review · 4 major / 5 minor · reviewed 2026-08-12 · deepseek-v4-flash
Pith's one-line read This paper develops a collocation method for the first boundary value problem of second-order linear differential equations, using incomplete even and odd trigonometric splines as approximating functions.
desk verdict Odd-spline collocation as printed is underdetermined, but the paired-spline method is plausible and worth a revision round. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The central objects are the fundamental even and odd trigonometric splines defined in equations (8)-(10) as infinite series with convergence factors σ(r-1, j). They are said to interpolate the grid values and to belong to $C^{{r-1}}$[0, π], which lets them be differentiated enough times to be substituted into a second-order equation. The spline interpolants (11)-(13) turn boundary conditions into direct assignments to the endpoint parameters, and collocation reduces the problem to a square linear algebraic system for the interior parameters.
What would settle it
Build the collocation matrix (19) for a small test problem, say Example 3 with r=3 and N=2, and test whether the matrix is invertible and whether each fundamental spline takes value 1 at its own node and 0 at the others; if the matrix is singular or the interpolation condition fails, the construction has no solution.
Extended reading notes
Core claim
On the paper's own terms, the central claim is that approximate solutions of the first boundary value problem (14)-(15) can be constructed as linear combinations of the fundamental even and odd trigonometric splines from earlier work [8], with nodal solution values as parameters. Three constructions are given: even splines on the grid Δ2_N^(0), odd splines on Δ3_N^(0), and odd splines on Δ3_N^(1). Boundary conditions fix some parameters directly; the remaining ones are found from the collocation systems (19) or (21). The numerical examples show that these approximate solutions stay close to the exact solutions, with reported maximum errors of about 0.048, 0.044, 0.042, 0.048, 0.045, 0.043, 0.00157, 0.00105, 0.00098, 0.0005, and 0.00063 depending on the example and spline order.
Load-bearing premise
The method stands on the unproved assertion, inherited from the authors' earlier paper, that the fundamental even and odd trigonometric splines really do interpolate the chosen grid values and are smooth enough to differentiate as required.
Editorial extensions
If this is right
- A user can apply the method to any linear second-order boundary value problem by mapping [a,b] to [0,π]; the endpoint conditions become the first and last spline parameters.
- Even splines should be used when the solution's first derivative vanishes at both endpoints, since the construction automatically enforces that extra condition.
- Odd splines are recommended for zero boundary conditions, because they vanish at 0 and π by construction.
- The error varies with the spline order r, so the examples indicate that order can be tuned to improve accuracy, though the paper does not give a general rule.
Reading between the lines
- If the interpolation property of these fundamental splines holds generally, the same collocation framework could be extended to higher-order differential equations or to systems by using tensor products of spline bases.
- The reported errors suggest that odd splines on zero-boundary problems may be significantly more accurate than even splines; a systematic comparison on the same problem would make this quantitative.
- A convergence proof, with error estimates as N grows for fixed r, is a natural next step; the paper only reports numerical behavior.
- For oscillatory solutions, trigonometric splines might need fewer degrees of freedom than polynomial splines, but the paper does not test this.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper proposes a collocation method for the first boundary value problem for second-order linear ODEs, using 'incomplete' even and odd fundamental trigonometric splines introduced in the authors' earlier work [8]. The approximate solution is written as a linear combination of fundamental splines (Eqs. 16-18), the coefficients are determined by collocation at prescribed interior grid points, and the resulting linear systems are given as Eqs. (19) and (21). Three numerical examples are presented: Examples 1 and 2 use even ('paired') splines, and Example 3 uses odd splines, with reported maximum deviations of order 0.04-0.05 for Examples 1-2 and 0.0005-0.0016 for Example 3. The paper concludes that the method is developed and that the error depends on the spline order r, but it provides no convergence proof, no nonsingularity analysis for the collocation systems, and a number of technical details are imported directly from reference [8].
Significance. If the method works as intended, it would be a modest but useful extension of spline-collocation techniques to non-polynomial (trigonometric) spline bases, potentially beneficial for problems with oscillatory or periodic features. The paper's main positive aspects are the clear conceptual framework (using fundamental splines so that boundary conditions can be imposed directly on coefficients) and the inclusion of three worked examples with explicit numerical errors. However, the current manuscript does not provide the rigor expected for a numerical-analysis paper: there is no theorem stating convergence or even well-posedness of the algebraic systems, and the supporting evidence rests entirely on a few smooth test problems. The most serious problems are that the odd-spline collocation system (21) is underdetermined as written and the exact solution in Example 3 does not satisfy the stated boundary value problem, so the central claim for method (b) is not yet supported. The paper would also benefit from reporting all discretization parameters (grid size, truncation of infinite series) so the examples can be reproduced.
major comments (4)
- [Eq. (21)] The system (21) is written for j=2,...,N-1, i.e., N-2 equations for N unknowns alpha_1,...,alpha_N. Since the odd fundamental splines vanish at x=0 and x=pi, the boundary conditions (20) are satisfied identically and do not determine any alpha_k. The text states that the values alpha_k for k=1,...,N are found, which would require N equations. If the intended range is j=1,...,N (all nodes of the grid (0)_3N, which are all interior), then the displayed range is a typo; otherwise method (b) is underdetermined and admits infinitely many approximate solutions. As printed, Example 3 cannot be reproduced from the stated algorithm, so this is a load-bearing defect in the central claim for method (b).
- [Example 3] In Example 3, the stated exact solution u(x) = -sin(x)/sin(1) satisfies u''+u = 0, not u''+u = -x as required by the displayed differential equation. Moreover, if the parenthetical domain is [0,1], then u(1) = -1, which contradicts the boundary condition u(1)=0; if the domain is [0,pi], the boundary condition u(pi)=0 holds but the equation is still not satisfied. Therefore the reported maximum deviations (0.00157, 0.00105, 0.00098, 0.0005, 0.00063) compare the approximate solution to a function that is not a solution of the stated boundary value problem. Example 3 must be corrected before it can support the odd-spline method.
- [Equations (8)-(10) and (11)-(13)] The fundamental even and odd spline formulas are quoted from reference [8] without proof or independent verification, and the claims that these functions satisfy the Kronecker-delta interpolation property and belong to C^{r-1}[0,pi] are asserted rather than demonstrated. Every entry of the collocation systems (19) and (21) depends on these properties, so their validity is load-bearing. The paper should either provide a self-contained proof (or at least a precise statement of the relevant theorem from [8]) or clearly state these properties as assumptions together with a specific reference to the theorem.
- [Examples 1-2] For none of the examples does the paper report the grid size N or the truncation parameters used when evaluating the infinite series in (8)-(10). Without this information, the reported maximum deviations (0.048, 0.044, 0.042 for Example 1; 0.048, 0.045, 0.043 for Example 2) are not reproducible from the manuscript alone. In addition, no condition numbers or convergence rates are given, so it is unclear whether the examples are representative of the method's behavior.
minor comments (5)
- [Eq. (16)] The index range 'k=2,2,...,N-1' should read 'k=2,3,...,N-1'.
- [Equations (8)-(10) and (16)-(18)] The independent variable is denoted t in the spline formulas (8)-(10) but x in the approximate solution (16)-(18); the domain of t and x should be stated consistently.
- [Figure 3] The text introducing the splines (10) concerns the grid (1)_3N, but the caption of Figure 3 says the vertical lines coincide with the nodes of the grid (0)_3N; this inconsistency should be corrected.
- [Abstract] The abstract states that 'theoretical positions are illustrated by numerical examples', but no theorem is stated; a precise convergence or error-bound statement would be more informative.
- [General typesetting] Several displayed equations (e.g., (8)-(10)) contain OCR-related artifacts and misaligned summation limits, which make them difficult to follow; a careful typesetting revision is needed.
Circularity Check
No circularity: coefficients are fixed by collocation and errors are checked against external exact solutions; the underdetermined Eq. (21) and reliance on the authors' prior spline formulas are correctness/reproducibility issues, not circularity.
full rationale
The predictive claims are validated against exact solutions of independent test problems, not against the inputs used to define the method. The approximate solution u*(x) is a linear combination of fundamental splines, and the coefficients are determined by collocation conditions L u*(x_j)=f(x_j), with boundary values imposed either through endpoint coefficients (case a) or automatically by the odd-spline basis (case b). Nothing in the derivation fits the reported maximum deviations to the exact solutions; those deviations are measured after the linear system is solved, so the accuracy assessment is not circular. The only self-citation is the use of the even/odd fundamental spline formulas from the authors' earlier work [8], and that is a standard, non-circular dependency: the current paper applies those splines to a new boundary-value-problem method rather than deriving the validity of the splines from the target examples. Two non-circular weaknesses should be noted as correctness/reproducibility risks, not circularity: (i) for the odd-spline case, system (21) is written for j=2,...,N-1, giving N-2 equations for N unknowns, so Example 3 cannot be reproduced from the method as stated; (ii) the examples report r but not the grid size N. These do not make the derivation circular because they concern determinacy and reproducibility, not prediction-by-construction.
Assumptions & free parameters
free parameters (1)
- spline order r =
3, 4, 5 (chosen in examples)
assumptions (4)
- domain assumption The fundamental even and odd trigonometric spline formulas from reference [8] are correct and have the stated interpolation and C^{r-1} properties.
- domain assumption The exact solution of the boundary value problem is sufficiently smooth and unique on [0, pi].
- domain assumption The collocation systems (19) and (21) have a unique solution for the chosen grids.
- standard math A linear change of variables maps a general interval [a,b] to [0,pi] while preserving the form of the boundary value problem.
Cite this review
Pith. "Pith review of Incomplete (even and odd) trigonometric splines in the problems of constructing approximate solutions of second order linear differential equations." pith.science (2026). https://pith.science/paper/EKAETU2N
@misc{pith2026241113524,
author = {Pith},
title = {Pith review of: Incomplete (even and odd) trigonometric splines in the problems of constructing approximate solutions of second order linear differential equations},
year = {2026},
howpublished = {\url{https://pith.science/paper/EKAETU2N}},
note = {Machine review of arXiv:2411.13524}
}
read the original abstract
The method of constructing approximate solutions of the first boundary value problem for linear differential equations based on incomplete (even and odd) trigonometric splines is considered. The theoretical positions are illustrated by numerical examples.
Figures
Reference graph
Works this paper leans on
-
[7]
Polynomial and trigonometric splines
Denysiuk V. Polynomial and trigonometric splines. arXiv:2110.04781
-
[8]
Forms of representation of interpolation trigonometric splines
Denysiuk V., Rybachuk L. Forms of representation of interpolation trigonometric splines. arXiv:2407.19275
-
[1]
Berezin I.S., Zhidkov N.P. Computing Methods, Vol. I,2. Pergamon Press, 1965 – 1143 p
work page 1965
-
[2]
Differential Equations and the Calculus of Variations
Elsgolts S. Differential Equations and the Calculus of Variations . University Press of the Pacific,
-
[3]
Computational Galerkin Methods // Springer-Verlag New York Berlin Heidelberg Tokyo, 1984
Fletcher C.A.J. Computational Galerkin Methods // Springer-Verlag New York Berlin Heidelberg Tokyo, 1984
work page 1984
-
[4]
The Theory of Splines and Their Applications
Ahlberg J.Н., Nilson E.N., Walsh J.L. The Theory of Splines and Their Applications. – Academic Press. New York/London, 1967
work page 1967
-
[5]
A survey of spline collocation methods for the numerical soluti on of differential equations
Fairweather G., Meade D. A survey of spline collocation methods for the numerical soluti on of differential equations. – CRC Press, Mathematics for large scale computing, 2020
work page 2020
-
[6]
Generalized Trigonometric Functions and Their Applications // IOSR Journal of Mathematics (IOSR-JM)
Denysiuk V. Generalized Trigonometric Functions and Their Applications // IOSR Journal of Mathematics (IOSR-JM). Volume 14, Issue 6 Ver. I, 2018, pp.19-25
work page 2018
Show all 10 references
-
[9]
S., Kudryavtsev L.D., Levitan B.M
Guter R. S., Kudryavtsev L.D., Levitan B.M. Elements of the Theory of Functions . – Translation edited by Sneddon I.N. – Pergamon Press, Oxford, 1966
1966
-
[10]
, Miroshnichenko V.L
Zavyalov Yu .S., Kvasov B.I. , Miroshnichenko V.L. Spline -function methods . Moskva: Nauka, 1980. 350 p. (Rus)
1980
Reviewed August 12, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.