Time Reparametrization, Not Fractional Calculus: A Reassessment of the Conformable Derivative
Aziz El Ghazouani, Fouad Ibrahim Abdou Amir, Khoulane Mohamed, M'hamed Elomari
TL;DR
This work analyzes the conformable derivative, arguing it is not a true fractional operator but a nonlinear time reparametrization of classical differentiation. By establishing that $D^{\\alpha}_{\\psi} f(t)=\\psi_{\\alpha}(t) f'(t)$ and introducing the time transform $\\tau=\\phi_{\\alpha}(t)$, the authors show that conformable models are mathematically equivalent to classical ones, yielding memory-free dynamics with altered time scales. Through ODE, PDE, evolution, and dynamical-systems examples, they demonstrate that conformable results can be recovered by a coordinate change, while Caputo/Riemann-Liouville derivatives produce genuine nonlocal memory effects. The findings urge researchers to treat conformable calculus as a computational reparametrization tool, not as a source of memory modeling, and to rely on established fractional calculus for memory-dependent phenomena. This has practical implications for modeling, analysis, and terminology in fractional-dynamics research.
Abstract
The conformable derivative has been promoted in numerous publications as a new fractional derivative operator. This article provides a critical reassessment of this claim. We demonstrate that the conformable derivative is not a fractional operator but a useful computational tool for systems with power-law time scaling, equivalent to classical differentiation under a nonlinear time reparametrization. Several results presented in the literature as novel fractional contributions can be reinterpreted within a classical framework. We show that problems formulated using the conformable derivative can be transformed into classical formulations via a change of variable. The solution is derived classically and then transformed back, this reformulation highlights the absence of genuinely nonlocal fractional effects. We provide a theoretical analysis, numerical simulations comparing conformable, classical, and truly fractional (Caputo) models, and discuss the reasons why this misconception persists. Our results suggest that classical derivatives, as well as established fractional derivatives, offer a more faithful framework for modeling memory-dependent phenomena.
