Pith. sign in

REVIEW 4 major objections 7 minor 39 references

M\"untz Pseudo Spectral Method: Theory and Numerical Experiments

T0 review · 4 major / 7 minor · reviewed 2026-08-14 · deepseek-v4-flash

Pith's one-line read The paper introduces two Lagrange–Müntz basis families, proves their interpolants achieve weighted error $O(N^{-m})$ for functions with endpoint singularities, and derives explicit Erdélyi–Kober fractional differentiation matrices in two…

desk verdict New Lagrange-Müntz bases with plausible theory and numerics, but the interpolation-space definitions don't match the basis spans and the paper leans on an unverified companion preprint. read the letter →

arxiv 1908.02306 v2 pith:CUCKZAZX submitted 2019-08-06 math.NA cs.NA

classification math.NAcs.NA MSC 26A3333C4541A5534L1065M7058C40
keywords Lagrange–MüntzbasisfunctionsJacobi–MüntzErdélyi–Koberfractionalderivativesdifferentiationmatricespseudo-spectralmethodnon-classicalinterpolantsMüntzquadraturerulesendpointsingularities
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

Two new Lagrange-type basis functions are introduced, both built from the authors' Jacobi–Müntz functions by multiplying a mapped cardinal function with a weight factor. The paper proves that the associated interpolants approximate any function in the appropriate mapped weighted space at rate $O(N^{-m})$, and it derives explicit left- and right-sided Erdélyi–Kober fractional differentiation matrices in two different ways. If the proofs are right, one parameterized family of bases covers the four classical Lagrange basis types and gives spectral-like accuracy for solutions with endpoint singularities, including fractional ordinary and partial differential equations. The second, matrix-based variant of the differentiation matrices is reported to remain accurate at much larger $N$ than the direct expansion formula.

What carries the argument

The load-bearing object is the pair of Lagrange–Müntz basis families: the mapped cardinal function $h_r^{\sigma}$ (the product over all $j\ne r$ of the mapped-node ratios) is multiplied, respectively, by a power of $x/x_r$ and by a power of $x/x_r$ times a power of $(b^\sigma-x^\sigma)$, so that each basis function carries the same weight structure as the Jacobi–Müntz functions it generalizes. Four supporting pieces carry the argument: the Gauss–Jacobi–Müntz quadrature rules make discrete inner products exact on the appropriate spans; the Jacobi–Müntz differentiation identity converts a fractional derivative of a basis element into another Jacobi–Müntz function with shifted parameters; the mapped-Jacobi interpolation error estimates provide the $O(N^{-m})$ rates; and the stable differentiation matrices are assembled as $U V^{-1}$, replacing a direct expansion that loses accuracy at large $N$.

What would settle it

Two concrete checks settle the matter: compare the exponents in the span definitions against the space $P_N$ used in Definition 3.9 for one parameter set, and compute a known Erdélyi–Kober derivative both through the matrix of Theorem 3.20 and through direct high-precision quadrature of the integral definition; a span mismatch or a disagreement beyond roundoff would show the claim as written is wrong.

Watch

Extended reading notes

Core claim

The paper's central claim is that the first basis function $(x/x_r)^{\sigma(\beta-\eta-\mu)}h_r^{\sigma}(x)$ and the second basis function $(x/x_r)^{\sigma\eta}((b^\sigma-x^\sigma)/(b^\sigma-x_r^\sigma))^{\alpha}h_r^{\sigma}(x)$ are genuine generalizations of the four known Lagrange basis families, where $h_r^{\sigma}$ is the cardinal product over mapped nodes. Theorems 3.12 and 3.17 establish stability and weighted error bounds of size $O(N^{-m})$ for the two non-classical Jacobi–Müntz interpolants. Theorems 3.18–3.21 then deliver the Erdélyi–Kober fractional differentiation matrices in two forms: one by expanding the cardinal functions and applying the Jacobi–Müntz differentiation identity term by term, and one as the product $U V^{-1}$ of the matrix of fractional derivatives of the Jacobi–Müntz basis and the inverse of its collocation matrix. The numerical experiments apply these matrices to linear, nonlinear, and partial fractional equations, and report that the stable variant keeps errors bounded with condition numbers growing like $O(N^{2\mu})$.

Load-bearing premise

The construction rests on the companion paper's formulas for differentiating and integrating Jacobi–Müntz functions holding in the parameter ranges used here, including restrictions such as $\beta-\mu>-1$ and $\alpha-\mu>-1$ that this paper does not restate or verify; the written identification of the interpolation space with $P_N$ must also match the actual span of the new bases.

Editorial extensions

If this is right

  • Fractional equations whose solutions have endpoint singularities can be collocated directly in these bases, and the paper proves weighted errors of size $O(N^{-m})$ for solutions in the appropriate mapped weighted spaces.
  • The parameter choices recover the four standard Lagrange basis families, so the new bases are a common umbrella for existing Lagrange-collocation schemes rather than a separate method.
  • Linear multi-term fractional equations reduce to a single dense linear system built from the differentiation matrices; nonlinear and time-dependent problems reduce to systems solved by Newton iteration or standard ODE solvers.
  • In the tested examples the stable matrix variant of the differentiation matrices remains accurate at substantially larger $N$ than the direct expansion formula, with condition numbers growing like $O(N^{2\mu})$.
  • For integer-order problems the same construction yields first-order differentiation matrices, and the Burgers example indicates that choosing $\sigma=1/2$ approximates singular solutions far better than the smooth case $\sigma=1$.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • This construction should extend to Gauss–Radau and Gauss–Lobatto nodes, which the paper mentions in passing; that would let the collocation method handle boundary and initial conditions without eliminating unknowns by hand.
  • One could automate parameter selection by estimating the leading singular exponent of the solution from the data and then choosing $\alpha,\beta,\eta,\mu,\sigma$ to match it, making the method an adaptive spectral scheme.
  • Composing the fractional differentiation matrices for several orders may give a route to variable- or distributed-order Erdélyi–Kober equations, although no composition rule is proved in the paper.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 7 minor

Summary. The paper introduces two new families of non-classical Lagrange basis functions (LMFs-1 and LMFs-2, Eq. (3.1)) built from Jacobi-Müntz functions, defines two associated interpolants, proves weighted error bounds of order O(N^{-m}) (Theorems 3.12, 3.16 and 3.17), and derives left- and right-sided Erdélyi-Kober fractional differentiation matrices both by direct nodal expansion (Theorems 3.18–3.19) and by a collocation-matrix inversion approach (Theorems 3.20–3.21 and 3.23). The numerical section applies the bases to fractional ODEs and PDEs, including a nonlinear Burgers example with endpoint singularities. The theoretical development relies on orthogonality, fractional derivative and quadrature results imported without proof from the companion preprint [13].

Significance. If the technical results hold, the construction unifies and generalizes the four Lagrange basis types in Table 1 and provides a spectral method tailored to endpoint singularities and Erdélyi-Kober fractional operators. The explicit closed form for the inverse of the collocation matrices in Theorem 3.23 is a practically useful contribution, and the numerical experiments in Section 4 cover a broad range of problems, with Examples 4.1–4.2 providing direct verification against exact fractional derivatives. The significance is conditional, however, on the correctness of the companion preprint [13] and on repairing the space-mismatch in Definition 3.9; the central claims are plausible but not yet fully rigorous as written.

major comments (4)
  1. [Section 3, Definition 3.9] Definition 3.9 declares the two interpolants to take values in P_N^{(β,μ,σ,η)} and P_N^{(α,σ,η)} as defined in (2.15)–(2.16), but those spaces contain exponents 2σ(β−η−μ), 2ση and 2α, whereas the LMFs in (3.1) and the spaces 1F_N and 2F_N in (3.26)–(3.27) contain exponents σ(β−η−μ), ση and α. Consequently the interpolants built from the LMFs are not elements of the stated codomains, and the discrete orthogonality identities immediately following Definition 3.9 are not justified as written. The factor 2 is correct for the product spaces used in the quadrature exactness statements (3.4), so the fix is to introduce separate notation for the interpolation spaces (e.g., 1F_N and 2F_N) and to use it consistently in Definition 3.9, in (3.38)–(3.39), and in Theorems 3.12 and 3.17.
  2. [Section 4, Example 4.5] The reduction of Equation (4.19) to the Riccati equation (4.20) is incorrect. For μ=1, Remark 2.4 gives aD^1_{x,σ,η} y = (1/σ)x y' + (η+1)y; substituting into (4.19) and multiplying by σ/x yields y' − 2y/x^2 = 1 − y^2, not y' − 2y = 1 − y^2. Hence the exact solution (4.21) does not solve the stated equation at μ=1, and the convergence reported in Figure 7 does not validate the method for this example. The equation, the exact solution, or both need to be corrected.
  3. [Sections 2.1–3.1] Theorems 3.18–3.21 apply the Jacobi-Müntz differentiation formulas of Remark 2.10 and the Gauss-Jacobi-Müntz quadrature rules of Theorem 2.13, all imported from the companion preprint [13] without proof or restatement of the required parameter restrictions. In particular, formula (3.40) contains Γ(j+β−μ+1) in the denominator and (3.48) contains Γ(j+α−μ+1), so the restrictions β−μ>−1 and α−μ>−1 are needed to avoid poles; these conditions are not stated in the theorems and are not checked in the numerical experiments. The paper should either state and verify these hypotheses explicitly or include proofs of the imported formulas in an appendix, since the differentiation matrices are a central claim of the paper.
  4. [Section 3.1, Theorems 3.20–3.21] The matrices L_SD^μ and R_SD^μ are defined as LU LV^{-1} and RU RV^{-1}, but no proof is given that the collocation matrices LV and RV are nonsingular for the stated parameters and nodes. Since the 'stable' differentiation matrices are obtained by inverting these dense matrices, the paper should prove, or at least state with a proof sketch, that the LMF basis is unisolvent at the Gauss-Jacobi-Müntz nodes. The numerical evidence in Figures 1–2 shows that even the second approach breaks down for sufficiently large N, so the terminology 'stable' in Theorems 3.20–3.21 should also be calibrated accordingly.
minor comments (7)
  1. [Section 4.2.1, Eq. (4.5)] The approximation y_N(x) is written as a sum over s=0,...,N of y(x_s) 1L_k(x); the index in the basis function should be s rather than k.
  2. [Equations (3.23) and (3.37)] The second interpolant is denoted with the superscript (α,β,μ,σ,η) in (3.23) and (3.37); the superscript should be (α,β,σ,η) to match Definition 3.9 and the right-sided basis in (3.1).
  3. [Table 1] In the row for Type 1, the entry g(x)=1 cannot be correct, because the cardinal basis in (1.4) would have denominator zero; the intended value is presumably g(x)=x.
  4. [Section 4.2.2, Example 4.6] The parameter line 'η = −µ = 1.75' is self-contradictory, since it implies μ=−1.75 and violates the stated condition 1<μ<2; it should read η=−1.75, μ=1.75.
  5. [Section 4.1, Example 4.2] The text refers to the 'left-sided EK fractional derivative' in Example 4.2, but the example uses the right-sided basis 2J and the right-sided matrices of Theorems 3.19 and 3.21; the wording should be 'right-sided'.
  6. [Remark 3.13] The displayed equality N^{-1}‖x^{-σ(β−η−μ)−1}u‖_{w(α+1,β+1,σ)} = N^{-1}‖x^{σ−2}(b^σ−x^σ)u‖_{w1^σ} has exponents that differ by x^2; the power σ−2 appears to be a misprint for σ−1 based on the preceding expressions.
  7. [References] References [29] and [30] are duplicates of the same paper, and reference [26] is missing author/editor information; these should be cleaned up.

Circularity Check

2 steps flagged · score 4.0 of 10

The O(N^{-m}) interpolation bounds are independently grounded in mapped-Jacobi theory, but the new fractional differentiation matrices and their numerical validation rest on unproved companion-preprint identities by the same authors, so the matrix claims are partly circular; a separate codomain mismatch in Definition 3.9 is a correctness gap rather than circularity.

  1. self citation load bearing [Section 2.1-2.2 (Remarks 2.9, 2.10, Theorem 2.13), used in Theorems 3.18-3.21 and (3.3)-(3.4)]
    "Proof. The proof of this theorem is presented in [13]. ... Proof. See [13] for the proof of this theorem."

    The fractional differentiation matrices in Theorems 3.18-3.21 are derived from the JMF derivative formula in Remark 2.10, and the discrete orthogonality (3.3)-(3.4) is justified by the Gauss-Jacobi-Muntz quadrature exactness of Theorem 2.13. Both results are taken without proof from the authors' own companion preprint [13] and are not verified by an independent, machine-checked, or externally reproducible argument in the present paper. Thus the validity of the new differentiation matrices reduces to a self-citation chain: if [13] contains an error or a missing parameter restriction such as beta-mu > -1, the present matrix formulas inherit that flaw.

  2. self definitional [Example 4.1, Eqs. (4.2) vs (3.40) and (3.52)]
    "As the first example consider f(x) = 1J (α,β,µ,σ,η) k (x). Using Remark 2.10 we arrive at: (4.2) 0Dµ x,σ,η [f(x)] = Γ(k +β + 1)/Γ(k +β−µ + 1) 1J (α+µ,β−µ,µ,σ,η−µ) k (x)"

    The matrix entries in Theorem 3.18 are produced by differentiating the JMF expansion with exactly the same Remark 2.10 identity, yielding the same Gamma ratio as in (3.40); Theorem 3.20 uses the same ratio in (3.52). Example 4.1 then measures the error of those differentiation matrices against the analytic derivative supplied by that same identity. Therefore the numerical 'validation' is a consistency check: the reference solution and the discretization share the same construction input, so the experiment can detect implementation or roundoff errors but cannot independently confirm the fractional-derivative formula or the matrix entries.

full rationale

The paper's central convergence claims are not circular. The new interpolants are factored through mapped-Jacobi interpolants in Remarks 3.10 and 3.15, and the O(N^{-m}) error bounds in Theorems 3.12, 3.16, and 3.17 follow from the standard mapped-Jacobi projection and interpolation estimates of Theorem 3.6, which are cited to the independent textbook Shen-Tang-Wang [24]. The Kronecker-delta nodal basis construction in (3.1)-(3.2) is direct and does not presuppose the target error bounds. The circularity exposure is concentrated in the fractional differentiation matrices and their numerical tests. Theorems 3.18-3.21 import the JMF derivative identities and the Gauss-Jacobi-Muntz quadrature exactness from the authors' own companion preprint [13] without proof here, and Examples 4.1-4.2 validate those matrices against the identical Remark 2.10 formula used in their derivation. This is a partial circularity: the matrix formulas reduce, in part, to self-cited and self-checked inputs, while the interpolation error analysis retains independent content. Separately, Definition 3.9 states that the interpolants map into P_N as defined in (2.15)-(2.16), whose exponents are 2σ(beta-mu-eta)+kσ, whereas the LMFs in (3.1) have leading exponent σ(beta-eta-mu)+kσ (and analogously for the second kind); as written, the stated codomain and the 'obvious' discrete orthogonality are not justified. This is an internal consistency/correctness gap, not a circularity, and I do not count it in the circularity score beyond noting it in the headline.

Assumptions & free parameters 5 free parameters · 5 assumptions · 0 invented entities

The method introduces no new physical entities. The central derivations rely on unproved results from the authors' companion paper [13], which is the main external-dependency risk; the error analysis otherwise sits on standard mapped Jacobi theory from [24].

free parameters (5)
  • alpha = 0.5, -0.5, and others per experiment
    Exponent in the weight (b^sigma - x^sigma)^alpha of the second basis family; chosen by hand to shape the singularity handling.
  • beta = 1, 2, 3 in examples
    Jacobi parameter controlling the mapped polynomial degree; chosen by hand.
  • sigma = 0.5, 1 in examples
    Müntz exponent; chosen to match power-type singularities such as sqrt(x).
  • eta = 0, -1, -2, -1.75 in examples
    Exponent in the weight factors of the basis; chosen by hand.
  • mu = 0.25, 0.5, 0.75, 0.95, 1, 1.75 in examples
    Fractional order of the derivative, a problem parameter, but it also appears in the LMFs-1 basis and is tuned in tests.
assumptions (5)
  • ad hoc to paper Jacobi-Müntz functions satisfy orthogonality (Remark 2.7) and completeness (Theorem 2.8)
    Stated as results from the authors' companion preprint [13]; not proved or independently verified in this paper.
  • ad hoc to paper Fractional EK derivatives of JMFs follow the closed forms in Remark 2.9/2.10 with shifted parameters
    These formulas are load-bearing for the differentiation matrices (Theorems 3.18 to 3.21) and are cited to [13]; parameter restrictions such as beta - mu > -1 and alpha - mu > -1 are not restated.
  • ad hoc to paper Gauss-Jacobi-Müntz quadrature rules (Theorem 2.13) are exact on the stated spaces
    Used to derive coefficients of the differentiation matrices; proof cited to [13].
  • standard math Mapped-Jacobi interpolation and projection error bounds (Theorem 3.6) from reference [24]
    Standard Jacobi approximation results, cited to the textbook of Shen, Tang and Wang.
  • ad hoc to paper The interpolation space in Definition 3.9 is the span of the new basis functions
    As written, Definition 3.9 uses the space P_N from (2.15) to (2.16), which has different exponents from the actual span (3.26) to (3.27); the paper silently assumes a corrected definition.

how reviews work

0 comments
Cite this review

Pith. "Pith review of M\"untz Pseudo Spectral Method: Theory and Numerical Experiments." pith.science (2026). https://pith.science/paper/CUCKZAZX

@misc{pith2026190802306,
  author       = {Pith},
  title        = {Pith review of: M\"untz Pseudo Spectral Method: Theory and Numerical Experiments},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/CUCKZAZX}},
  note         = {Machine review of arXiv:1908.02306}
}
read the original abstract

This paper presents two new non-classical Lagrange basis functions which are based on the new Jacobi-M\"untz functions presented by the authors recently. These basis functions are, in fact, generalizations form of the newly generated Jacobi based functions. With respect to these non-classical Lagrange basis functions, two non-classical interpolants are introduced and their error bounds are proved in detail. The pseudo-spectral differentiation (and integration) matrices have been extracted in two different manners. Some numerical experiments are provided to show the efficiency and capability of these newly generated non-classical Lagrange basis functions.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

39 extracted references · 33 canonical work pages

  1. [13]

    Esmaeili, M

    S. Esmaeili, M. Shamsi, and Y. Luchko, Numerical solution of fractional dif- ferential equations with a collocation method based on M¨ untz polynomials, Com- puters & Mathematics with Applications. An International Journal, 62 (2011), pp. 918–929

  2. [1]

    M\"untz Pseudo Spectral Method: Theory and Numerical Experiments

    Introduction. The history of fractional calculus goes back to 17th century. In fact, the fractional calculus deals with the calculus of the integrals and derivatives of non-integer (real or complex) orders. So, the fractional calculus can be considered as a generalization of the classical calculus [19, 21, 26]. Up to now, several definitions of fractional ...

  3. [2]

    In this section, we compile some basic definitions and prop- erties of fractional differential operators

    Preliminaries. In this section, we compile some basic definitions and prop- erties of fractional differential operators. Definition 2.1. The left and right Erd´ elyi-Kober fractional integralsaIµ x,σ,η and xIµ b,σ,η of order µ∈ R+ are defined by [14]: (2.1) aIµ x,σ,η[f](x) = σx−σ(η+µ) Γ(µ) ∫ x a (xσ−tσ)µ−1tσ(η+1)−1f(t)dt, x∈ (a,b ], a> 0, and (2.2) xIµ b,σ,η...

  4. [3]

    This section devotes to the main results of this paper

    Main results. This section devotes to the main results of this paper. To do so, we introduce two new non-classical Lagrange basis functions corresponding to the newly introduced basis functions JMFs-1 and JMFs-2 as follows: Definition 3.1. Let{xr}N r=0 be an arbitrary set of nodes on [0,b ], then the Lagrange-M¨ untz basis functions of the first- and secon...

  5. [4]

    This section is concerned to testify the theoretical results numerically

    Numerical experiments. This section is concerned to testify the theoretical results numerically. To do so, we divide this section into two parts. In the first part, applications of the newly interpolants to approximate the EK fractional derivatives are given. In the second part, applications of these interpolants to solve some ordinary and fractional parti...

  6. [5]

    This paper presents two new non-classical Lagrange basis functions which are, in fact, generalizations of all the previous Lagrange basis functions

    Concluding remarks. This paper presents two new non-classical Lagrange basis functions which are, in fact, generalizations of all the previous Lagrange basis functions. Theoretical results with respect to these basis functions are developed in detail. Some numerical experiments are provided to verify the theoretical results. Some future works are listed b...

  7. [6]

    T. M. Atanackovi´c, S. Pilipovi ´c, B. Stankovi ´c, and D. Zorica , Frac- tional calculus with applications in mechanics , Mechanical Engineering and Solid Mechanics Series, ISTE, London; John Wiley & Sons, Inc., Hoboken, NJ, 2014

  8. [7]

    Baleanu, K

    D. Baleanu, K. Diethelm, E. Scalas, and J. J. Trujillo , Fractional calculus, vol. 3 of Series on Complexity, Nonlinearity and Chaos, World Scien- tific Publishing Co. Pte. Ltd., Hackensack, NJ, 2012, https://doi.org/10.1142/ 9789814355216

Show all 39 references
  1. [8]

    J. P. Boyd, Chebyshev and Fourier spectral methods , Dover Publications, Inc., Mineola, NY, second ed., 2001

  2. [9]

    Canuto, M

    C. Canuto, M. Y. Hussaini, A. Quarteroni, and T. A. Zang , Spectral methods, Scientific Computation, Springer, Berlin, 2007

  3. [10]

    T. S. Chihara, An introduction to orthogonal polynomials, Gordon and Breach Science Publishers, New York-London-Paris, 1978

  4. [11]

    Ejlali and S

    N. Ejlali and S. M. Hosseini, A pseudospectral method for fractional optimal control problems, Journal of Optimization Theory and Applications, 174 (2017), pp. 83–107, https://doi.org/10.1007/s10957-016-0936-8

  5. [12]

    Esmaeili and M

    S. Esmaeili and M. Shamsi , A pseudo-spectral scheme for the approximate solution of a family of fractional differential equations , Communications in Nonlinear Science and Numerical Simulation, 16 (2011), pp. 3646–3654, https: //doi.org/10.1016/j.cnsns.2010.12.008

  6. [14]

    J. S. Hesthaven, S. Gottlieb, and D. Gottlieb , Spectral methods for time-dependent problems , vol. 21 of Cambridge Monographs on Applied and Computational Mathematics, Cambridge University Press, Cambridge, 2007, https://doi.org/10.1017/CBO9780511618352

  7. [15]

    Hilfer, Applications of fractional calculus in physics , World Scientific Pub- This manuscript is for review purposes only

    R. Hilfer, Applications of fractional calculus in physics , World Scientific Pub- This manuscript is for review purposes only. M¨UNTZ PSEUDO–SPECTRAL METHOD: THEORY AND NUMERICAL EXPERIMENTS 29 lishing Co., Inc., River Edge, NJ, 2000, https://doi.org/10.1142/9789812817747

  8. [16]

    Huang, Y

    C. Huang, Y. Jiao, L.-L. Wang, and Z. Zhang , Optimal fractional inte- gration preconditioning and error analysis of fractional collocation method us- ing nodal generalized Jacobi functions , SIAM Journal on Numerical Analysis, 54 (2016), pp. 3357–3387, https://doi.org/10.1137...

  9. [17]

    Kharazmi, M

    E. Kharazmi, M. Zayernouri, and G. E. Karniadakis , A Petrov-Galerkin spectral element method for fractional elliptic problems , Computer Methods in Applied Mechanics and Engineering, 324 (2017), pp. 512–536, https://doi.org/ 10.1016/j.cma.2017.06.006

  10. [18]

    Khosravian-Arab and M

    H. Khosravian-Arab and M. R. Eslahchi , M¨ untz Sturm-Liouville prob- lems: Theory and numerical experiments , 2019, https://arxiv.org/abs/arXiv: 1908.00062

  11. [19]

    A. A. Kilbas, H. M. Srivastava, and J. J. Trujillo , Theory and applica- tions of fractional differential equations , vol. 204 of North-Holland Mathematics Studies, Elsevier Science B.V., Amsterdam, 2006

  12. [20]

    Li and F

    C. Li and F. Zeng , Numerical methods for fractional calculus , Chapman & Hall/CRC Numerical Analysis and Scientific Computing, CRC Press, Boca Ra- ton, FL, 2015

  13. [21]

    Mainardi , Fractional calculus and waves in linear viscoelasticity , Imperial College Press, London, 2010, https://doi.org/10.1142/9781848163300

    F. Mainardi , Fractional calculus and waves in linear viscoelasticity , Imperial College Press, London, 2010, https://doi.org/10.1142/9781848163300

  14. [22]

    Mao and G

    Z. Mao and G. E. Karniadakis , A spectral method (of exponential conver- gence) for singular solutions of the diffusion equation with general two-sided frac- tional derivative, SIAM Journal on Numerical Analysis, 56 (2018), pp. 24–49

  15. [23]

    M. M. Meerschaert and A. Sikorskii , Stochastic models for fractional cal- culus, vol. 43 of De Gruyter Studies in Mathematics, Walter de Gruyter & Co., Berlin, 2012

  16. [24]

    K. S. Miller and B. Ross , An introduction to the fractional calculus and fractional differential equations, A Wiley-Interscience Publication, John Wiley & Sons, Inc., New York, 1993

  17. [25]

    M. D. Ortigueira , Fractional calculus for scientists and engineers , vol. 84 of Lecture Notes in Electrical Engineering, Springer, Dordrecht, 2011, https: //doi.org/10.1007/978-94-007-0747-4

  18. [26]

    Podlubny, Fractional differential equations, vol

    I. Podlubny, Fractional differential equations, vol. 198 of Mathematics in Sci- ence and Engineering, Academic Press, Inc., San Diego, CA, 1999

  19. [27]

    Sabatier, O

    J. Sabatier, O. P. Agrawal, and J. A. T. Machado, Advances in fractional calculus, Springer, Dordrecht, 2007, https://doi.org/10.1007/978-1-4020-6042-7

  20. [28]

    Saha Ray, Fractional calculus with applications for nuclear reactor dynamics, CRC Press, Boca Raton, FL, 2016

    S. Saha Ray, Fractional calculus with applications for nuclear reactor dynamics, CRC Press, Boca Raton, FL, 2016

  21. [29]

    J. Shen, T. Tang, and L.-L. Wang , Spectral methods, vol. 41 of Springer Series in Computational Mathematics, Springer, Heidelberg, 2011, https://doi. org/10.1007/978-3-540-71041-7

  22. [30]

    J. L. Suzuki, M. Zayernouri, M. L. Bittencourt, and G. E. Karni- adakis, Fractional-order uniaxial visco-elasto-plastic models for structural anal- ysis, Computer Methods in Applied Mechanics and Engineering, 308 (2016), pp. 443–467, https://doi.org/10.1016/j.cma.2016.05.030

  23. [31]

    Fractional order calculus and its applications , Springer, Dordrecht, 2002

  24. [32]

    L. N. Trefethen , Spectral methods in MATLAB , vol. 10 of Software, Envi- ronments, and Tools, Society for Industrial and Applied Mathematics (SIAM), Philadelphia, PA, 2000, https://doi.org/10.1137/1.9780898719598

  25. [33]

    J. A. C. Weideman and L. N. Trefethen , The eigenvalues of second-order This manuscript is for review purposes only. 30 HASSAN KHOSRAVIAN-ARAB AND MOHAMMAD REZA ESLAHCHI spectral differentiation matrices, SIAM Journal on Numerical Analysis, 25 (1988), pp. 1279–1298, https://doi...

  26. [34]

    Zayernouri, M

    M. Zayernouri, M. Ainsworth, and G. Karniadakis , A unified petrov- galerkin spectral method for fractional pdes , Computer Methods in Applied Me- chanics and Engineering, 283 (2015), pp. 1545–1569, https://doi.org/10.1016/j. cma.2014.10.051

  27. [35]

    Zayernouri, M

    M. Zayernouri, M. Ainsworth, and G. E. Karniadakis, A unified Petrov- Galerkin spectral method for fractional PDEs , Computer Methods in Applied Mechanics and Engineering, 283 (2015), pp. 1545–1569, https://doi.org/10.1016/ j.cma.2014.10.051

  28. [36]

    Zayernouri and G

    M. Zayernouri and G. E. Karniadakis , Fractional spectral collocation method, SIAM Journal on Scientific Computing, 36 (2014), pp. A40–A62, https: //doi.org/10.1137/130933216

  29. [37]

    Zayernouri and G

    M. Zayernouri and G. E. Karniadakis, Fractional spectral collocation meth- ods for linear and nonlinear variable order FPDEs , Journal of Computational Physics, 293 (2015), pp. 312–338, https://doi.org/10.1016/j.jcp.2014.12.001

  30. [38]

    Zayernouri, S.-W

    M. Zayernouri, S.-W. Park, D. M. Tartakovsky, and G. E. Karni- adakis, Stochastic smoothed profile method for modeling random roughness in flow problems, Computer Methods in Applied Mechanics and Engineering, 263 (2013), pp. 99–112, https://doi.org/10.1016/j.cma.2013.05.007

  31. [39]

    F. Zeng, Z. Mao, and G. E. Karniadakis , A generalized spectral collocation method with tunable accuracy for fractional differential equations with end-point singularities, SIAM Journal on Scientific Computing, 39 (2017), pp. A360–A383, https://doi.org/10.1137/16M1076083. This m...

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.