Part of a series of articles about
Calculus
${\displaystyle f(a)-f(b)=\int _{b}^{a}f'(t)\,dt}$
Fundamental theorem Limits Continuity Rolle's theorem Mean value theorem Inverse function theorem
Differential Definitions Derivative (generalizations) Differential infinitesimal of a function total Concepts Differentiation notation Second derivative Implicit differentiation Logarithmic differentiation Related rates Taylor's theorem Rules and identities Sum Product Chain Power Quotient L'Hôpital's rule Inverse General Leibniz Faà di Bruno's formula Reynolds
Integral Lists of integrals Integral transform Leibniz integral rule Definitions Antiderivative Integral (improper) Riemann integral Lebesgue integration Contour integration Integral of inverse functions Integration by Parts Discs Cylindrical shells Substitution (trigonometric, tangent half-angle, Euler) Euler's formula Partial fractions Changing order Reduction formulae Differentiating under the integral sign Risch algorithm
Series Geometric (arithmetico-geometric) Harmonic Alternating Power Binomial Taylor Convergence tests Summand limit (term test) Ratio Root Integral Direct comparison Limit comparison Alternating series Cauchy condensation Dirichlet Abel
Vector Gradient Divergence Curl Laplacian Directional derivative Identities Theorems Gradient Green's Stokes' Divergence generalized Stokes
Multivariable Formalisms Matrix Tensor Exterior Geometric Definitions Partial derivative Multiple integral Line integral Surface integral Volume integral Jacobian Hessian
Advanced Calculus on Euclidean space Generalized functions Limit of distributions
Specialized Fractional Malliavin Stochastic Variations
Miscellaneous Precalculus History Glossary List of topics Integration Bee Mathematical analysis Nonstandard analysis
v t e

In mathematics, the Gateaux differential or Gateaux derivative is a generalization of the concept of directional derivative in differential calculus. Named after René Gateaux, a French mathematician who died at age 25 in World War I, it is defined for functions between locally convex topological vector spaces such as Banach spaces. Like the Fréchet derivative on a Banach space, the Gateaux differential is often used to formalize the functional derivative commonly used in the calculus of variations and physics.

Unlike other forms of derivatives, the Gateaux differential of a function may be nonlinear. However, often the definition of the Gateaux differential also requires that it be a continuous linear transformation. Some authors, such as Tikhomirov (2001), draw a further distinction between the Gateaux differential (which may be nonlinear) and the Gateaux derivative (which they take to be linear). In most applications, continuous linearity follows from some more primitive condition which is natural to the particular setting, such as imposing complex differentiability in the context of infinite dimensional holomorphy or continuous differentiability in nonlinear analysis.

Definition

Suppose ${\displaystyle X}$ and ${\displaystyle Y}$ are locally convex topological vector spaces (for example, Banach spaces), ${\displaystyle U\subseteq X}$ is open, and ${\displaystyle F:U\to Y.}$ The Gateaux differential ${\displaystyle dF(u;\psi )}$ of ${\displaystyle F}$ at ${\displaystyle u\in U}$ in the direction ${\displaystyle \psi \in X}$ is defined as

${\displaystyle dF(u;\psi )=\lim _{\tau \to 0}{\frac {F(u+\tau \psi )-F(u)}{\tau ))=\left.{\frac {d}{d\tau ))F(u+\tau \psi )\right|_{\tau =0))$

(1)

If the limit exists for all ${\displaystyle \psi \in X,}$ then one says that ${\displaystyle F}$ is Gateaux differentiable at ${\displaystyle u.}$

The limit appearing in (1) is taken relative to the topology of ${\displaystyle Y.}$ If ${\displaystyle X}$ and ${\displaystyle Y}$ are real topological vector spaces, then the limit is taken for real ${\displaystyle \tau .}$ On the other hand, if ${\displaystyle X}$ and ${\displaystyle Y}$ are complex topological vector spaces, then the limit above is usually taken as ${\displaystyle \tau \to 0}$ in the complex plane as in the definition of complex differentiability. In some cases, a weak limit is taken instead of a strong limit, which leads to the notion of a weak Gateaux derivative.

Linearity and continuity

At each point ${\displaystyle u\in U,}$ the Gateaux differential defines a function

$${\displaystyle dF(u;\cdot ):X\to Y.}$$

This function is homogeneous in the sense that for all scalars ${\displaystyle \alpha ,}$

$${\displaystyle dF(u;\alpha \psi )=\alpha dF(u;\psi ).\,}$$

However, this function need not be additive, so that the Gateaux differential may fail to be linear, unlike the Fréchet derivative. Even if linear, it may fail to depend continuously on ${\displaystyle \psi }$ if ${\displaystyle X}$ and ${\displaystyle Y}$ are infinite dimensional. Furthermore, for Gateaux differentials that are linear and continuous in ${\displaystyle \psi ,}$ there are several inequivalent ways to formulate their continuous differentiability.

For example, consider the real-valued function ${\displaystyle F}$ of two real variables defined by

$${\displaystyle F(x,y)={\begin{cases}{\dfrac {x^{3)){x^{2}+y^{2))}&{\text{if ))(x,y)\neq (0,0),\\0&{\text{if ))(x,y)=(0,0).\end{cases))}$$

${\displaystyle (0,0).}$

$${\displaystyle dF(0,0;a,b)={\begin{cases}{\dfrac {F(\tau a,\tau b)-0}{\tau ))&(a,b)\neq (0,0),\\0&(a,b)=(0,0)\end{cases))={\begin{cases}{\dfrac {a^{3)){a^{2}+b^{2))}&(a,b)\neq (0,0),\\0&(a,b)=(0,0).\end{cases))}$$

${\displaystyle (a,b).}$

${\displaystyle X}$

${\displaystyle 0}$

Relation with the Fréchet derivative

If ${\displaystyle F}$ is Fréchet differentiable, then it is also Gateaux differentiable, and its Fréchet and Gateaux derivatives agree. The converse is clearly not true, since the Gateaux derivative may fail to be linear or continuous. In fact, it is even possible for the Gateaux derivative to be linear and continuous but for the Fréchet derivative to fail to exist.

Nevertheless, for functions ${\displaystyle F}$ from a complex Banach space ${\displaystyle X}$ to another complex Banach space ${\displaystyle Y,}$ the Gateaux derivative (where the limit is taken over complex ${\displaystyle \tau }$ tending to zero as in the definition of complex differentiability) is automatically linear, a theorem of Zorn (1945). Furthermore, if ${\displaystyle F}$ is (complex) Gateaux differentiable at each ${\displaystyle u\in U}$ with derivative

$${\displaystyle DF(u):\psi \mapsto dF(u;\psi )}$$

${\displaystyle F}$

${\displaystyle U}$

${\displaystyle DF}$

Continuous differentiability

Continuous Gateaux differentiability may be defined in two inequivalent ways. Suppose that ${\displaystyle F:U\to Y}$ is Gateaux differentiable at each point of the open set ${\displaystyle U.}$ One notion of continuous differentiability in ${\displaystyle U}$ requires that the mapping on the product space

$${\displaystyle dF:U\times X\to Y\,}$$

${\displaystyle X}$

${\displaystyle Y}$

${\displaystyle dF(u;\cdot )}$

${\displaystyle u}$

A stronger notion of continuous differentiability requires that

$${\displaystyle u\mapsto DF(u)\,}$$

$${\displaystyle U\to L(X,Y)\,}$$

${\displaystyle U}$

${\displaystyle X}$

${\displaystyle Y.}$

${\displaystyle DF(u).}$

As a matter of technical convenience, this latter notion of continuous differentiability is typical (but not universal) when the spaces ${\displaystyle X}$ and ${\displaystyle Y}$ are Banach, since ${\displaystyle L(X,Y)}$ is also Banach and standard results from functional analysis can then be employed. The former is the more common definition in areas of nonlinear analysis where the function spaces involved are not necessarily Banach spaces. For instance, differentiation in Fréchet spaces has applications such as the Nash–Moser inverse function theorem in which the function spaces of interest often consist of smooth functions on a manifold.

Higher derivatives

Whereas higher order Fréchet derivatives are naturally defined as multilinear functions by iteration, using the isomorphisms ${\displaystyle L^{n}(X,Y)=L(X,L^{n-1}(X,Y)),}$ higher order Gateaux derivative cannot be defined in this way. Instead the ${\displaystyle n}$th order Gateaux derivative of a function ${\displaystyle F:U\subseteq X\to Y}$ in the direction ${\displaystyle h}$ is defined by

${\displaystyle d^{n}F(u;h)=\left.{\frac {d^{n)){d\tau ^{n))}F(u+\tau h)\right|_{\tau =0}.}$

(2)

Rather than a multilinear function, this is instead a homogeneous function of degree ${\displaystyle n}$ in ${\displaystyle h.}$

There is another candidate for the definition of the higher order derivative, the function

${\displaystyle D^{2}F(u)\{h,k\}=\lim _{\tau \to 0}{\frac {DF(u+\tau k)h-DF(u)h}{\tau ))=\left.{\frac {\partial ^{2)){\partial \tau \,\partial \sigma ))F(u+\sigma h+\tau k)\right|_{\tau =\sigma =0))$

(3)

that arises naturally in the calculus of variations as the second variation of ${\displaystyle F,}$ at least in the special case where ${\displaystyle F}$ is scalar-valued. However, this may fail to have any reasonable properties at all, aside from being separately homogeneous in ${\displaystyle h}$ and ${\displaystyle k.}$ It is desirable to have sufficient conditions in place to ensure that ${\displaystyle D^{2}F(u)\{h,k\))$ is a symmetric bilinear function of ${\displaystyle h}$ and ${\displaystyle k,}$ and that it agrees with the polarization of ${\displaystyle d^{n}F.}$

For instance, the following sufficient condition holds (Hamilton 1982). Suppose that ${\displaystyle F}$ is ${\displaystyle C^{1))$ in the sense that the mapping

$${\displaystyle DF:U\times X\to Y}$$

$${\displaystyle D^{2}F:U\times X\times X\to Y}$$

${\displaystyle D^{2}F(u)\{h,k\))$

${\displaystyle h}$

${\displaystyle k.}$

$${\displaystyle D^{2}F(u)\{h,k\}={\frac {1}{2))d^{2}F(u;h+k)-d^{2}F(u;h)-d^{2}F(u;k)}$$

${\displaystyle D^{2}F(u)}$

${\displaystyle d^{2}F(u;-).}$

Properties

A version of the fundamental theorem of calculus holds for the Gateaux derivative of ${\displaystyle F,}$ provided ${\displaystyle F}$ is assumed to be sufficiently continuously differentiable. Specifically:

• Suppose that ${\displaystyle F:X\to Y}$ is ${\displaystyle C^{1))$ in the sense that the Gateaux derivative is a continuous function ${\displaystyle dF:U\times X\to Y.}$ Then for any ${\displaystyle u\in U}$ and ${\displaystyle h\in X,}$
  $${\displaystyle F(u+h)-F(u)=\int _{0}^{1}dF(u+th;h)\,dt}$$

  
  where the integral is the Gelfand–Pettis integral (the weak integral) (Vainberg (1964)).

Many of the other familiar properties of the derivative follow from this, such as multilinearity and commutativity of the higher-order derivatives. Further properties, also consequences of the fundamental theorem, include:

• (The chain rule)
  
  $${\displaystyle d(G\circ F)(u;x)=dG(F(u);dF(u;x))}$$

  
  for all ${\displaystyle u\in U}$ and ${\displaystyle x\in X.}$ (Importantly, as with simple partial derivatives, the Gateaux derivative does not satisfy the chain rule if the derivative is permitted to be discontinuous.)

• (Taylor's theorem with remainder)
  Suppose that the line segment between ${\displaystyle u\in U}$ and ${\displaystyle u+h}$ lies entirely within ${\displaystyle U.}$ If ${\displaystyle F}$ is ${\displaystyle C^{k))$ then
  $${\displaystyle F(u+h)=F(u)+dF(u;h)+{\frac {1}{2!))d^{2}F(u;h)+\dots +{\frac {1}{(k-1)!))d^{k-1}F(u;h)+R_{k))$$

  
  where the remainder term is given by
  $${\displaystyle R_{k}(u;h)={\frac {1}{(k-1)!))\int _{0}^{1}(1-t)^{k-1}d^{k}F(u+th;h)\,dt}$$

  

Example

Let ${\displaystyle X}$ be the Hilbert space of square-integrable functions on a Lebesgue measurable set ${\displaystyle \Omega }$ in the Euclidean space ${\displaystyle \mathbb {R} ^{n}.}$ The functional

$${\displaystyle E:X\to \mathbb {R} }$$

$${\displaystyle E(u)=\int _{\Omega }F(u(x))\,dx}$$

${\displaystyle F}$

${\displaystyle u}$

${\displaystyle \Omega }$

$${\displaystyle dE(u;\psi )=\langle F'(u),\psi \rangle :=\int _{\Omega }F'(u(x))\,\psi (x)\,dx.}$$

Indeed, the above is the limit ${\displaystyle \tau \to 0}$ of

$${\displaystyle {\begin{aligned}{\frac {E(u+\tau \psi )-E(u)}{\tau ))&={\frac {1}{\tau ))\left(\int _{\Omega }F(u+\tau \,\psi )\,dx-\int _{\Omega }F(u)\,dx\right)\\[6pt]&={\frac {1}{\tau ))\left(\int _{\Omega }\int _{0}^{1}{\frac {d}{ds))F(u+s\,\tau \,\psi )\,ds\,dx\right)\\[6pt]&=\int _{\Omega }\int _{0}^{1}F'(u+s\tau \psi )\,\psi \,ds\,dx.\end{aligned))}$$