Fact-checked by Grok 2 weeks ago

Legendre transformation

The Legendre transformation, named after the French mathematician Adrien-Marie Legendre (1752–1833), is a fundamental operation in mathematics that converts a convex function f: \mathbb{R}^n \to \mathbb{R} into its convex conjugate or dual function f^*: (\mathbb{R}^n)^* \to \mathbb{R}, defined as f^*(p) = \sup_{x \in \mathbb{R}^n} (p \cdot x - f(x)), where p belongs to the dual space and represents the gradient or slope associated with x.^[1] This transform effectively encodes the supporting hyperplanes of the function's epigraph, switching the natural variables from the original domain to their conjugate momenta or derivatives while preserving the function's convexity and essential information.^[2] Introduced by Legendre in 1787 in the context of solving partial differential equations related to minimal surfaces and variational problems, the transformation gained prominence through its applications in classical mechanics and thermodynamics, where it facilitates changes between conjugate pairs of variables such as position and momentum or energy and temperature.^[3] For a differentiable convex function f(x), the transform simplifies to f^*(p) = p x(p) - f(x(p)), with p = f'(x) and the inverse relation x = (f^*)'(p), ensuring the operation is involutive—applying it twice yields the original function.^[1] Key properties include the symmetry f(x) + f^*(p) = p \cdot x for conjugate pairs (x, p), and the inverse relationship between second derivatives, \frac{d^2 f^*}{dp^2} = \left( \frac{d^2 f}{dx^2} \right)^{-1}, which highlights its role in optimizing curvatures and dual formulations.^[2] In classical mechanics, it underpins the transition from the Lagrangian L(q, \dot{q}) to the Hamiltonian H(q, p) = p \dot{q} - L(q, \dot{q}) with p = \partial L / \partial \dot{q}, enabling phase-space analysis of dynamical systems.^[1] Similarly, in thermodynamics and statistical mechanics, it relates potentials like internal energy U(S, V) to enthalpy H(S, P) = U + P V or Helmholtz free energy F(T, V) = U - T S, where conjugate variables such as entropy S and temperature T = \partial U / \partial S are interchanged to suit different experimental conditions.^[2] Beyond physics, the Legendre transformation is central to convex analysis and optimization, where it duality allows reformulating constrained problems as unconstrained suprema, and in calculus of variations, it aids in deriving Euler-Lagrange equations for functionals involving convex integrands.^[1] Its geometric interpretation as the convex hull's envelope of tangent planes underscores its utility in fields like contact geometry and machine learning for convex risk minimization.^[3]

Definition

One-dimensional case

The Legendre transformation of a convex function f: \mathbb{R} \to \mathbb{R}, also known as the convex conjugate or Fenchel conjugate, is defined by the formula

f^*(p) = \sup_{x \in \mathbb{R}} \left( p x - f(x) \right),

where p \in \mathbb{R} serves as the dual variable representing the slope in the geometric interpretation.^[4]^[5] This supremum captures the maximum value of the linear approximation p x minus f(x), ensuring f^* is convex as the pointwise supremum of affine functions.^[4] Applying the transformation twice yields the biconjugate

f^{**}(x) = \sup_{p \in \mathbb{R}} \left( p x - f^*(p) \right).

For any convex lower semicontinuous function f, the biconjugate recovers the original function exactly, i.e., f^{**} = f, which underscores the involutive nature of the operation under these conditions.^[4]^[5] This recovery property relies on the lower semicontinuity to close the convex envelope of f.^[4] When f is strictly convex and differentiable, the supremum in the definition of f^* is uniquely attained at the point x satisfying p = f'(x), and the inverse relation holds with x = (f^*)'(p).^[4] In this case, the functions satisfy the Young-Fenchel equality f(x) + f^*(p) = p x, providing a direct duality between the primal and dual variables.^[4]^[5] The transformation is named after the French mathematician Adrien-Marie Legendre (1752–1833), who introduced it in 1787 in the context of solving partial differential equations related to minimal surfaces and variational problems. The modern involutive and duality aspects were formalized shortly thereafter by Pierre-Simon Laplace in his 1788 work on curve generation.

Finite-dimensional case

In the finite-dimensional case, the Legendre transformation, also known as the Legendre-Fenchel conjugate, extends to functions f: \mathbb{R}^n \to \mathbb{R} defined on the Euclidean space \mathbb{R}^n. For such a function, the Legendre transform f^*: \mathbb{R}^n \to \mathbb{R} is given by

f^*(p) = \sup_{x \in \mathbb{R}^n} \left( p \cdot x - f(x) \right),

where p, x \in \mathbb{R}^n and \cdot denotes the standard dot product.^[5] This formulation generalizes the one-dimensional case by replacing the scalar product px with the vector dot product p \cdot x.^[5] The supremum in the definition of f^*(p) is attained at a point x satisfying \nabla f(x) = p, provided f is differentiable at x.^[6] For the maximizer to be unique, f must be strictly convex, ensuring that the gradient \nabla f is injective and the supporting hyperplane touches the graph of f at exactly one point.^[5] Under these conditions, the Legendre transform establishes a bijection between the domain of f and the domain of f^*, with the inverse mapping recovered via the gradient: \nabla f^*(p) = x where x solves \nabla f(x) = p.^[6] A key property is the biconjugacy theorem, which states that if f is convex and lower semicontinuous, then the double Legendre transform recovers the original function: f^{**} = f.^[5] This involution holds because the conjugate f^* encodes the supporting hyperplanes of the convex epigraph of f, and applying the transform twice reconstructs the envelope of those hyperplanes.^[5] If f is twice continuously differentiable with a positive definite Hessian \nabla^2 f(x) > 0, then f^* is also twice differentiable, and the Hessians satisfy (\nabla^2 f(x))^{-1} = \nabla^2 f^*(p) at corresponding points where \nabla f(x) = p.^[6] This inverse relationship reflects the duality in the curvature of f and f^*, preserving the strict convexity. A canonical example is the quadratic function f(x) = \frac{1}{2} \|x\|^2, where \| \cdot \|^2 is the squared Euclidean norm. The supremum \sup_x (p \cdot x - \frac{1}{2} \|x\|^2) is attained at x = p, yielding f^*(p) = \frac{1}{2} \|p\|^2.^[5] Here, f^* = f, illustrating self-duality for this isotropic quadratic form, with \nabla^2 f(x) = I (the identity matrix) and thus \nabla^2 f^*(p) = I as well.^[6]

Interpretation via derivatives

The Legendre transform can be interpreted as a change of variables from the original independent variable x to its conjugate variable p, defined as the slope of the tangent to the graph of f(x), given by p = \frac{df}{dx}.^[7] This substitution encodes the same information as f(x) but in terms of p, facilitating analysis when derivatives are more natural independent variables, such as in optimization or phase space formulations.^[2] The transform is invertible under suitable conditions, with the inverse relation x = \frac{d f^*}{dp}, where f^*(p) is the Legendre transform of f(x), reflecting the duality between the functions.^[8] To compute f^*(p) operationally, solve the equation p = f'(x) for x as a function of p, then substitute into the expression f^*(p) = p x - f(x).^[7] This algorithm assumes the derivative f'(x) is invertible, which requires f(x) to be strictly convex, satisfying f''(x) > 0 for all x, ensuring a bijective mapping between x and p.^[2] Without this condition, the mapping may not be one-to-one, leading to non-uniqueness in the inversion.^[8] From a second-order perspective, for twice-differentiable f(x), the second derivative of the transform relates inversely to that of the original function: \frac{d^2 f^*}{dp^2} = \left( \frac{d^2 f}{dx^2} \right)^{-1}, evaluated at corresponding points.^[9] This relation arises from differentiating the inversion x(p), and it connects to local approximations where the Taylor expansion of f(x) around a point uses quadratic terms governed by f''(x), mirroring how the transform approximates the function via its tangent lines in interpolation contexts.^[9] An integration-by-parts interpretation views the Legendre transform as shifting integrals from the x-domain to the p-domain: \int f'(x) \, dx = p x - \int \frac{d f^*}{dp} \, dp, which is particularly useful in asymptotic analysis, such as Laplace's method for evaluating integrals dominated by stationary points.^[9] Here, the transform simplifies the asymptotic expansion by aligning the integration variable with the slope at the extremum, enhancing computational efficiency for large-parameter limits.^[10]

Core Properties

Convexity and biconjugacy

Convexity plays a fundamental role in the Legendre transformation, ensuring its well-definedness, differentiability, and invertibility properties. For a non-convex function f, the Legendre transform f^* may fail to be differentiable or exhibit unique invertibility, as the supremum defining f^* might not yield a single point or smooth behavior. In contrast, if f is convex, then f^* is also convex, which guarantees that the mapping induced by the gradients (where they exist) is bijective under appropriate growth conditions.^[11] A key result establishing the recovery property of the transform is the Fenchel-Moreau theorem, which states that for a proper lower semicontinuous convex function f: \mathbb{R}^n \to (-\infty, +\infty], the biconjugate f^{**} = f. This theorem implies that applying the Legendre transform twice recovers the original function exactly within the class of proper lower semicontinuous convex functions, highlighting the involutive nature of the operation on this domain.^[11] The Young-Fenchel inequality underpins these duality relations, asserting that for any function f and its Legendre transform f^*,

f(x) + f^*(p) \geq p \cdot x \quad \forall x, p \in \mathbb{R}^n,

with equality holding if and only if p \in \partial f(x), where \partial f(x) denotes the subdifferential of f at x. This inequality provides a variational characterization of the duality, quantifying the gap between f(x) and the supporting hyperplane approximation at that point.^[11] The subdifferential further characterizes the attainment in the supremum defining f^*(p): the maximum in f^*(p) = \sup_x (p \cdot x - f(x)) is achieved at points x such that p \in \partial f(x). This extends the classical derivative condition to non-smooth convex functions, where the subdifferential replaces the gradient, ensuring the transform remains meaningful even without differentiability.^[11] Finally, the Legendre transform preserves the structure of closed convex lower semicontinuous functions: if f belongs to this class, so does f^*, and thus f^{**}. This closure property reinforces the transform's utility in convex analysis, maintaining the relevant regularity conditions throughout iterations.^[11]

Monotonicity and invertibility

For a convex function f: \mathbb{R}^n \to \mathbb{R}, the gradient mapping x \mapsto \nabla f(x) (where defined) is monotone, meaning that (\nabla f(x_1) - \nabla f(x_2)) \cdot (x_1 - x_2) \geq 0 for all x_1, x_2 in the interior of the domain of f. This property follows from the convexity of f, as the subdifferential at any point contains gradients that satisfy the monotonicity condition via the definition of subgradients. If f is strictly convex, the mapping is strictly monotone, with the inequality becoming strict for x_1 \neq x_2.^[4] The monotonicity of the gradient ensures desirable mapping properties for the Legendre transform f^*(p) = \sup_x \{ p \cdot x - f(x) \}. Specifically, if f is strictly convex and essentially smooth—meaning it is differentiable on the interior of its domain and |\nabla f(x)| \to \infty as x approaches the boundary of the domain—then the gradient \nabla f is a bijection from the interior of \dom f onto the interior of \dom f^*.^[12] Under these conditions, the Legendre transform establishes a bijection between the primal space (associated with f) and the dual space (associated with f^*), with the inverse mapping given by \nabla f^* = (\nabla f)^{-1}. This invertibility theorem highlights the transform's role in duality, allowing recovery of the original function through the dual gradient.^[4] A key consequence of this bijection is the self-inverse nature of the Legendre transform for convex functions. For a proper convex lower semicontinuous function f, applying the transform twice yields the biconjugate: (f^*)^* = f. When f is of Legendre type (strictly convex and essentially smooth), the transform is precisely self-inverse without needing closure, as the bijection ensures exact recovery.^[12] This involution property underscores the symmetry between f and f^*, facilitating iterative applications in optimization and analysis.^[4] The Legendre transform also exhibits compatibility with affine transformations of the variables. For an invertible linear map A \in \mathbb{R}^{n \times n}, if g(x) = f(Ax), then the conjugate satisfies g^*(p) = f^*(A^{-T} p), where A^{-T} = (A^{-1})^T. For general affine changes g(x) = f(Ax + b), the transform adjusts accordingly: g^*(p) = f^*(A^{-T} p) - b \cdot A^{-T} p, preserving the duality structure under these variable shifts. This commutation allows the transform to be applied consistently across linearly related coordinate systems.^[11] Finally, strict convexity guarantees the uniqueness of maximizers in the defining supremum of the Legendre transform. For each p in the interior of \dom f^*, there exists a unique x achieving \sup_x \{ p \cdot x - f(x) \}, given by x = \nabla f^*(p).^[4] This uniqueness stems from the strict monotonicity of \nabla f, ensuring a single point where the hyperplane tangent to the graph of f has slope p.

Geometric Interpretation

Envelopes and supporting hyperplanes

The geometric interpretation of the Legendre transformation provides insight into its role in convex analysis, particularly through the lens of envelopes formed by tangent lines and supporting hyperplanes. For a convex function f: \mathbb{R}^n \to \mathbb{R}, the Legendre transform f^*(p) = \sup_x (p \cdot x - f(x)) encodes the function via its supporting hyperplanes. Specifically, when f is differentiable, for each p = \nabla f(x) in the range of the gradient, f^*(p) equals the negative y-intercept (in the affine sense) of the tangent hyperplane to the graph of f at x, given by the equation y = p \cdot z - f^*(p). This hyperplane supports the epigraph of f from below at (x, f(x)), meaning the epigraph lies entirely above it.^[13] In the one-dimensional case, this construction is particularly visualizable. Consider a strictly convex, differentiable function f: \mathbb{R} \to \mathbb{R}. The tangent line at a point x_0 has slope p = f'(x_0) and equation y = p (z - x_0) + f(x_0) = p z - f^*(p), where f^*(p) = p x_0 - f(x_0). Plotting these tangent lines for varying x_0 (or equivalently, varying p), their lower envelope traces the graph of f itself, since f is convex and the tangents lie above the graph. Conversely, the graph of f^* emerges as the lower convex envelope of the family of lines y = x z - f(x) parametrized by x, illustrating the duality: points on the graph of f^* correspond to slopes of tangents to f, and vice versa. This envelope property preserves the information content of f, as the transform inverts via the biconjugate f^{**} = f.^[7]^[13] The relation to the convex hull underscores the transformation's convexifying effect. For a general (not necessarily convex) function f, the biconjugate f^{**} is the convex envelope of f, the greatest convex function majorized by f, obtained as the lower envelope of all supporting lines to the epigraph of f. In this context, the epigraph of f^* serves as the polar dual to the epigraph of f, capturing the set-theoretic duality where supporting hyperplanes to one correspond to points in the other. This duality links the Legendre transform to the support function of the (closure of the) convex hull of the set \{(x, -f(x)) \mid x \in \dom f\}.^[5] Parametrically, the boundary points of the graph of f^* satisfy the classical Legendre condition for envelopes. Consider the family of lines y(z) = f'(x) z - (f'(x) x - f(x)) indexed by the parameter x corresponding to points on f. The envelope condition requires \partial / \partial x [y - f'(x) z + f^*(f'(x))] = 0, yielding z = x (since (f^*)'(p) = x), and substituting back confirms the graph of f^* as the envelope, with points (p, f^*(p)) satisfying the dual relation x = (f^*)'(p). In higher dimensions, this generalizes to the condition that the supporting hyperplane with normal p touches the epigraph at x where p = \nabla f(x).^[14]

Duality in convex sets

The Legendre transform establishes a fundamental duality in convex analysis, linking a proper lower semicontinuous convex function f: \mathbb{R}^n \to (-\infty, \infty] to its conjugate f^*, defined as

f^*(p) = \sup_{x \in \mathbb{R}^n} \bigl( p \cdot x - f(x) \bigr).

This conjugate represents the support function of the epigraph of f, where the epigraph \mathrm{epi}(f) = \{ (x, t) \in \mathbb{R}^{n+1} \mid t \geq f(x) \} is a closed convex set. Specifically, evaluating the support function of \mathrm{epi}(f) in the direction (p, -1) yields f^*(p), providing a geometric view through supporting hyperplanes to the epigraph of f. This interpretation highlights how the transform encodes the boundary behavior of f via linear functionals. The polar transform extends this duality from functions to convex sets. For a nonempty convex set K \subseteq \mathbb{R}^n containing the origin, the polar set is

K^\circ = \{ p \in \mathbb{R}^n \mid p \cdot x \leq 1 \ \forall \, x \in K \},

which is itself closed and convex. This construction is analogous to the Legendre conjugate of the indicator function \delta_K(x) = 0 if x \in K and +\infty otherwise, whose conjugate is the support function h_K(p) = \sup_{x \in K} p \cdot x; thus, K^\circ = \{ p \mid h_K(p) \leq 1 \}. The polar duality preserves convexity and reverses inclusion order, mirroring the order-reversing property of the functional Legendre transform. Biduality for sets parallels biconjugacy for functions: the bidual (K^\circ)^\circ = \overline{\mathrm{conv}} \, K, the closure of the convex hull of K. If K is closed and convex with the origin in its interior, then (K^\circ)^\circ = K, ensuring the transform is an involution on this class of sets. This property underscores the symmetry in convex geometry, where polarity twice returns to the original set up to closure and convexification. In optimization, the Legendre transform facilitates dualization of convex programs through Lagrangian duality. For the problem \min_x f_0(x) subject to f_i(x) \leq 0 (i=1,\dots,m), the Lagrangian is L(x, \lambda) = f_0(x) + \sum_i \lambda_i f_i(x) with \lambda \succeq 0, and the dual function g(\lambda) = \inf_x L(x, \lambda) = - \sup_x \bigl( -L(x, \lambda) \bigr) involves the negative conjugate of an effective objective. The dual problem \max_{\lambda \succeq 0} g(\lambda) provides a lower bound, with strong duality holding under Slater's condition; for linear programs, this corresponds to the standard LP dual, interpretable via polarity of the feasible polyhedron. Gauge functions further connect the transform to duality in norms and sets. For a closed convex set K \subseteq \mathbb{R}^n with nonempty interior containing the origin, the gauge (Minkowski functional) is \|x\|_K = \inf \{ t > 0 \mid x \in t K \}, a convex 1-homogeneous function. Its Legendre conjugate is the indicator \delta_{K^\circ} of the polar set, since \|p\|_{K^\circ} = \sup_{x \in K} p \cdot x / \|x\|_K = \sup_{\|x\|_K \leq 1} p \cdot x = h_K(p) under homogeneity, yielding the dual gauge (or dual norm if K is symmetric) as the support function of K. Thus, \|\cdot\|_{K^\circ} = h_K, linking the original gauge to its polar counterpart. The Legendre transform is the unique order-reversing involution (up to affine adjustments) on the space of proper closed convex functions, characterizing it as the canonical duality operator in convex analysis.^[15]

Examples

Linear and quadratic functions

The Legendre transform of a linear function f(x) = c x + d in one dimension, where c, d \in \mathbb{R}, is computed as f^*(p) = \sup_x (p x - c x - d). This supremum equals -d if p = c and +\infty otherwise, corresponding to the indicator function of the point p = c shifted by the constant -d.^[13] In the special case where d = 0, the transform simplifies to f^*(p) = 0 if p = c and +\infty otherwise, which in the sense of distributions behaves like a Dirac delta function centered at p = c. More generally, affine adjustments to a base function affect the transform in a straightforward manner. For f(x) = g(x) + a x + b with a, b \in \mathbb{R}, the Legendre transform is f^*(p) = g^*(p - a) - b, where g^* is the transform of the base function g. This property allows constants and linear terms to be handled separately, preserving the structure of the original transform while applying translations and shifts.^[13] For quadratic functions, consider the multivariate case f(\mathbf{x}) = \frac{1}{2} \mathbf{x}^T A \mathbf{x} + \mathbf{b}^T \mathbf{x} + c, where A is a symmetric positive definite matrix (A \succ 0), \mathbf{b} \in \mathbb{R}^n, and c \in \mathbb{R}. The Legendre transform f^*(\mathbf{p}) is obtained by solving \mathbf{p} = \nabla f(\mathbf{x}) = A \mathbf{x} + \mathbf{b}, yielding \mathbf{x} = A^{-1} (\mathbf{p} - \mathbf{b}). Substituting gives

f^*(\mathbf{p}) = \frac{1}{2} (\mathbf{p} - \mathbf{b})^T A^{-1} (\mathbf{p} - \mathbf{b}) - c.

^[13] To verify in one dimension, take f(x) = \frac{1}{2} k x^2 with k > 0, so A = k, \mathbf{b} = 0, and c = 0. Then p = k x, x = p / k, and f^*(p) = p (p / k) - \frac{1}{2} k (p / k)^2 = p^2 / k - \frac{1}{2} p^2 / k = \frac{1}{2} p^2 / k. This illustrates the inverse scaling by the Hessian (second derivative k), a general feature for strictly convex quadratics.^[13] If the quadratic is not convex (e.g., A \not\succeq 0), the Legendre transform may become multi-valued or take values in the extended reals, as the supremum can fail to achieve a unique maximum or require handling non-convex branches via supporting hyperplanes. In such cases, the biconjugate f^{**} provides the convex envelope of f, effectively convexifying the function through the transform process.

Exponential and logarithmic functions

The Legendre transform of the exponential function f(x) = e^x is computed by finding the supremum \sup_x (p x - e^x). The critical point occurs where the derivative vanishes: p - e^x = 0, so x = \ln p for p > 0. Substituting yields f^*(p) = p \ln p - p for p > 0, f^*(0) = 0, and f^*(p) = +\infty for p < 0.^[9]^[13] This form highlights the transform's domain restriction to the nonnegative half-line, with a singularity at p = 0^+ where f^*(p) \to 0 from below. A shifted variant, f(x) = e^x - 1, follows similarly: p = e^x, x = \ln p, and f^*(p) = p \ln p - p + 1 for p > 0, f^*(0) = 1, and +\infty for p < 0.^[9]^[13] For the negative branch, considering the convex function f(x) = e^{-x} (which is the exponential reflected), the slope p = f'(x) = -e^{-x} < 0. Solving gives x = -\ln(-p), and substitution results in f^*(p) = -p \ln(-p) + p for p < 0, f^*(0) = 0, and +\infty for p > 0.^[16] The term -p \ln(-p) emerges as the core nonlinear part, defined on the negative half-line with a singularity at p = 0^- where it approaches $0 from below. The logarithmic function f(x) = -\ln x for x > 0 is strictly convex, with derivative p = f'(x) = -1/x < 0. Inverting yields x = -1/p, valid for p < 0. The transform is f^*(p) = \sup_{x>0} (p x + \ln x) = -1 - \ln(-p) for p < 0, and +\infty otherwise.^[16] Restricting the domain to $0 < x \leq 1 limits slopes to p \leq -1, so f^*(p) = -1 - \ln(-p) for p < -1, with boundary behavior at p = -1 where x=1 and the function value is finite, but diverging as p \to -\infty. This form relates to relative entropy measures, as the expression mirrors components of the Kullback-Leibler divergence for distributions on bounded supports. A self-dual pair arises in probability theory through the cumulant generating function and its transform. In one dimension, consider f(x) = \ln(\sum_i e^{x_i}) simplified to the exponential case f(x) = x (degenerate), but more relevantly, the pair f(x) = e^x - 1 and f^*(p) = p \ln p - p + 1 demonstrates biconjugacy, where applying the transform twice recovers the original up to the convex hull. In general, for cumulant functions like f(x) = \ln \mathbb{E}[e^{x Y}], the Legendre transform yields the rate function I(p) = \sup_x (p x - f(x)), which is self-dual in the sense of convex duality for exponential families. These transforms are inherently defined on half-lines due to the monotonicity of the original functions, with logarithmic singularities at the boundaries reflecting the asymptotic behavior of the originals near zero or infinity.

Applications in Physics

Thermodynamics and potentials

In thermodynamics, the Legendre transformation provides a systematic framework for defining various thermodynamic potentials by changing the independent variables from extensive quantities like entropy S and volume V to their intensive conjugates, such as temperature T and pressure p. The internal energy U(S, V) serves as the fundamental potential, representing the total energy as a function of S and V for a closed system.^[17] The enthalpy H is obtained via the Legendre transform of U with respect to V, given by H(S, p) = U(S, V) + p V, where p = -\left( \frac{\partial U}{\partial V} \right)_S is the conjugate pressure. This transform shifts the natural variables to S and p, making H suitable for processes at constant pressure, such as in open systems or calorimetric measurements. Similarly, the Helmholtz free energy F results from transforming U with respect to S: F(T, V) = U(S, V) - T S, with T = \left( \frac{\partial U}{\partial S} \right)_V, changing variables to T and V for isothermal conditions. The Gibbs free energy G combines both transforms: G(T, p) = U - T S + p V = F + p V = H - T S, with natural variables T and p, ideal for phase equilibria and chemical reactions at constant temperature and pressure.^[17]^[2]^[18] These potentials yield exact differential forms that facilitate thermodynamic relations. For instance, the differential of the Helmholtz free energy is dF = -S \, dT - p \, dV, from which Maxwell relations arise due to the equality of mixed second partial derivatives. A key example is \left( \frac{\partial S}{\partial V} \right)_T = \left( \frac{\partial p}{\partial T} \right)_V, linking thermal expansion and pressure changes. Similar relations follow from other potentials, such as \left( \frac{\partial S}{\partial p} \right)_T = -\left( \frac{\partial V}{\partial T} \right)_p from dG = -S \, dT + V \, dp. These relations enable computation of hard-to-measure properties from accessible ones, underpinning much of applied thermodynamics.^[18]^[19] The convexity of the internal energy U(S, V) in its natural variables ensures thermodynamic stability, as the positive definiteness of the second derivatives—\left( \frac{\partial T}{\partial S} \right)_V > 0 (specific heat) and -\left( \frac{\partial p}{\partial V} \right)_S > 0 (compressibility)—prevents unphysical fluctuations and guarantees equilibrium uniqueness. This convexity property, inherent to the Legendre framework, extends to the other potentials through their transforms, maintaining physical consistency across representations.^[17]^[2]

Analytical mechanics

In analytical mechanics, the Legendre transformation provides a fundamental method for reformulating the dynamics of a system from the Lagrangian description to the Hamiltonian description. The Lagrangian L(\mathbf{q}, \dot{\mathbf{q}}) is a function of generalized coordinates \mathbf{q} and their time derivatives \dot{\mathbf{q}}, typically expressed as the difference between kinetic and potential energies. To obtain the Hamiltonian, the Legendre transform is applied with respect to the velocities \dot{\mathbf{q}}, defining the conjugate momenta as p_i = \frac{\partial L}{\partial \dot{q}_i} for each component i. The Hamiltonian H(\mathbf{q}, \mathbf{p}) is then given by

H(\mathbf{q}, \mathbf{p}) = \sum_i \dot{q}_i p_i - L(\mathbf{q}, \dot{\mathbf{q}}),

where the velocities \dot{\mathbf{q}} are expressed as functions of \mathbf{q} and \mathbf{p} via the inverse of the momentum definition. This transformation shifts the description from the configuration-velocity space (\mathbf{q}, \dot{\mathbf{q}}) to the phase space (\mathbf{q}, \mathbf{p}), which is a $2n-dimensional manifold for a system with n degrees of freedom, enabling a symmetric treatment of positions and momenta.^[2]^[20] For standard mechanical systems where the kinetic energy is quadratic in the velocities, the Lagrangian is convex in \dot{\mathbf{q}}, ensuring that the Legendre transform is well-defined and invertible, and the resulting Hamiltonian is convex in the momenta \mathbf{p}. This convexity property guarantees a one-to-one correspondence between velocities and momenta, facilitating the analysis of stability and equilibrium in phase space. The equations of motion in the Hamiltonian formulation, known as Hamilton's equations,

\dot{q}_i = \frac{\partial H}{\partial p_i}, \quad \dot{p}_i = -\frac{\partial H}{\partial q_i},

emerge naturally from the duality of the Legendre transform, as the partial derivatives recover the original velocities and forces from the Lagrangian. These first-order equations provide an equivalent yet often more insightful description of the system's evolution compared to the second-order Euler-Lagrange equations.^[20]^[2] A simple example illustrates this transformation for a free particle in one dimension, where the Lagrangian is L = \frac{1}{2} m \dot{q}^2 with mass m. The momentum is p = \frac{\partial L}{\partial \dot{q}} = m \dot{q}, so \dot{q} = p/m. Substituting into the transform yields the Hamiltonian H = \frac{p^2}{2m}, which depends only on the momentum and describes the particle's kinetic energy in phase space. The corresponding Hamilton's equations are \dot{q} = p/m and \dot{p} = 0, reproducing the uniform motion \mathbf{q}(t) = \mathbf{q}_0 + (p/m) t.^[21]^[20] The Legendre transformation preserves the symplectic structure of the phase space, ensuring that the change from Lagrangian to Hamiltonian coordinates constitutes a canonical transformation. Such transformations maintain the Poisson bracket \{f, g\} = \sum_i \left( \frac{\partial f}{\partial q_i} \frac{\partial g}{\partial p_i} - \frac{\partial f}{\partial p_i} \frac{\partial g}{\partial q_i} \right) between any two functions f and g, which underpins the conservation laws and integrability of Hamiltonian systems via Noether's theorem and Liouville's theorem. This invariance allows for the generation of new conserved quantities and facilitates advanced techniques like action-angle variables.^[20]^[22]

Applications in Other Fields

Economics and optimization

In microeconomics, the Legendre transformation underpins the duality between production functions and cost functions, enabling the recovery of one from the other under appropriate convexity assumptions. The cost function c(\mathbf{w}, y) for input prices \mathbf{w} and output level y is defined as c(\mathbf{w}, y) = \inf_{\mathbf{x}} \{ \mathbf{w} \cdot \mathbf{x} \mid f(\mathbf{x}) \geq y \}, where f(\mathbf{x}) is the production function mapping inputs \mathbf{x} to output. For a concave and increasing production function f, this cost function is convex and increasing in y, and the original production function can be recovered via the dual relation f(\mathbf{x}) = \sup \{ y \mid c(\mathbf{w}, y) \leq \mathbf{w} \cdot \mathbf{x} \} for all \mathbf{w} > 0. This duality framework was formalized by Shephard, who established the theoretical foundations linking cost minimization to production technology.^[23] A key implication of this duality is Shephard's lemma, which states that the partial derivative of the cost function with respect to an input price equals the optimal input demand: \frac{\partial c(\mathbf{w}, y)}{\partial w_i} = x_i^*(\mathbf{w}, y). This result follows from the envelope theorem applied to the cost minimization problem, where changes in prices directly affect the value function through the optimal solution without needing to recompute second-order adjustments. Shephard's lemma facilitates empirical estimation of production technologies from observed cost data and has been central to advances in productivity analysis.^[24]^[23] On the consumer side, the Legendre transformation relates the utility function u(\mathbf{x}) to the indirect utility function v(\mathbf{p}, m) = \sup_{\mathbf{x}} \{ u(\mathbf{x}) \mid \mathbf{p} \cdot \mathbf{x} \leq m \}, where \mathbf{p} are prices and m is income. For quasi-concave and increasing utility, v is decreasing and quasi-convex in prices, and the duality allows recovery of preferences from expenditure or demand observations. The envelope theorem yields Roy's identity: x_i^*(\mathbf{p}, m) = -\frac{\partial v / \partial p_i}{\partial v / \partial m}, linking optimal demands to derivatives of the indirect utility. This transform structure supports welfare analysis and revealed preference tests.^[25]^[23] In profit maximization, the profit function \pi(\mathbf{p}, \mathbf{w}) = \sup_{\mathbf{x}} \{ \mathbf{p} \cdot f(\mathbf{x}) - \mathbf{w} \cdot \mathbf{x} \} serves as the Legendre transform of the cost function, with Hotelling's lemma stating \frac{\partial \pi(\mathbf{p}, \mathbf{w})}{\partial p_i} = y_i^*(\mathbf{p}, \mathbf{w}), the optimal output supply. Derived via the envelope theorem on the profit maximization problem, this lemma parallels Shephard's and enables derivation of supply functions from profit data, confirming the conjugate duality under the Legendre-Fenchel framework for convex technologies.^[24]^[23] In convex optimization, the Lagrangian dual function g(\boldsymbol{\lambda}) = \inf_{\mathbf{x}} L(\mathbf{x}, \boldsymbol{\lambda}), where L(\mathbf{x}, \boldsymbol{\lambda}) = f(\mathbf{x}) + \boldsymbol{\lambda} \cdot (b - A\mathbf{x}) is the Lagrangian for constraints A\mathbf{x} \leq b, relates to the Legendre transform through convex conjugation: g(\boldsymbol{\lambda}) = \mathbf{b} \cdot \boldsymbol{\lambda} - f^*(A^T \boldsymbol{\lambda}) for \boldsymbol{\lambda} \geq 0, where f^* is the convex conjugate of f. Under Slater's condition for strong duality, the primal and dual optima coincide, and the transform provides geometric insight into supporting hyperplanes for feasible sets. This connection has been instrumental in developing efficient algorithms like interior-point methods.^[26]^[13]^[23]

Probability and information theory

In probability theory, the Legendre transformation plays a central role in large deviations principles, particularly through the cumulant generating function. For a random variable X with moment generating function M(\theta) = \mathbb{E}[e^{\theta \cdot X}], the cumulant generating function is defined as K(\theta) = \log M(\theta), which is convex and generates the cumulants as its successive derivatives at the origin. The Legendre transform of K is the rate function I(x) = \sup_{\theta} (\theta \cdot x - K(\theta)), providing the exponential decay rate of rare events where the sample mean deviates from its expectation. This construction underlies Cramér's theorem, which states that for independent and identically distributed random variables X_i, the empirical mean S_n/n satisfies a large deviations principle with speed n and good rate function I.^[27] Sanov's theorem extends this framework to the space of empirical measures. For i.i.d. samples from a probability measure P on a Polish space, the empirical measure L_n = \frac{1}{n} \sum_{i=1}^n \delta_{X_i} obeys a large deviations principle with speed n and rate function I(\mu) = D(\mu \| P), the Kullback-Leibler divergence, which equals the Legendre transform of the logarithmic partition function \log \int e^{\theta} \, dP over suitable test functions \theta. This rate function quantifies the improbability of observing atypical empirical distributions, with the transform ensuring the principle holds in the weak topology.^[27] The relative entropy, or Kullback-Leibler divergence D(P \| Q) = \int \log \frac{dP}{dQ} \, dP, admits a variational representation as a Legendre-type transform: D(P \| Q) = \sup_f \left( \mathbb{E}_P - \log \mathbb{E}_Q[e^f] \right), where the supremum is over bounded continuous functions f. This form, known as the Donsker-Varadhan representation, highlights the divergence as the convex conjugate of the cumulant generating functional \log \mathbb{E}_Q[e^f] with respect to the reference measure Q, connecting it directly to large deviations rate functions. The duality between moment and cumulant generating functions further illustrates the transform's role. The moment generating function M(\theta) yields raw moments \mathbb{E}[X^k] as its k-th derivative at zero, while the cumulant generating function K(\theta) = \log M(\theta) provides cumulants \kappa_k similarly, with cumulants expressing moments via combinatorial relations and vice versa. In large deviations, the Legendre transform of K bridges these, yielding rate functions that decay exponentially for deviations, as the convexity of K ensures the transform is lower semicontinuous and non-negative.^[27] A variational perspective on entropy emerges through the Legendre transform in statistical physics contexts. The entropy S(\mu) = -\int \log \frac{d\mu}{d\pi} \, d\mu of a measure \mu absolutely continuous with respect to a reference \pi can be represented variationally as S(\mu) = -\inf_{\nu} \left( \mathbb{E}_\nu[H] + D(\nu \| \pi) \right), where H is the Hamiltonian; the infimum yields the free energy F(\beta) = -\frac{1}{\beta} \log Z(\beta) as the transform of the entropy, with \beta the inverse temperature. This links probabilistic rates to thermodynamic potentials, where the partition function Z(\beta) acts as the moment generating function for energy fluctuations.^[28]

Advanced Extensions

On manifolds and infinite dimensions

The Legendre transformation extends naturally to Riemannian manifolds, where the inner product is defined via the metric tensor g. For a function f on the tangent bundle TM of a Riemannian manifold (M, g), the transform is given by f^*(p) = \sup_{x \in TM} \langle p, x \rangle_g - f(x), with the duality pairing \langle p, x \rangle_g = g^{ij} p_i x_j incorporating the inverse metric components. This formulation establishes a duality between the tangent bundle TM and the cotangent bundle T^*M, where the Legendre map acts fiberwise as a symplectomorphism, preserving the canonical symplectic structure and relating Lagrangian mechanics on TM to Hamiltonian mechanics on T^*M.^[29] An analog of the Tonelli theorem applies to Lagrangians on compact manifolds, ensuring the existence of minimizing geodesics and complete flows. For a Tonelli Lagrangian L: TM \to \mathbb{R}, which is C^2, strictly convex in the velocity variable, and superlinear at infinity, the Legendre transform (x, v) \mapsto (x, \partial L / \partial v (x, v)) is a global diffeomorphism from TM to T^*M. This map conjugates the Euler-Lagrange flow on TM to the Hamiltonian flow on T^*M, with the associated Hamiltonian H(x, p) = \langle p, v \rangle_g - L(x, v) satisfying Tonelli conditions such as strict convexity and superlinearity. Tonelli minimizers—curves minimizing the action functional \int L(\gamma(t), \dot{\gamma}(t)) \, dt—are C^1 and satisfy the Euler-Lagrange equations, with their flows complete under compactness of M.^[30] In infinite-dimensional settings, the Legendre transform is defined on a Banach space X with dual X^* as f^*: X^* \to \mathbb{R}, f^*(p) = \sup_{x \in X} \langle p, x \rangle - f(x), for proper lower semicontinuous convex functions f: X \to \mathbb{R}. In Hilbert spaces, where reflexivity holds, the transform preserves key properties like biconjugacy (f^{**} = f) and induces isometries via Moreau-Yosida approximations. For general reflexive Banach spaces, similar results extend using Mosco epi-convergence, but compactness requires the weak* topology on bounded sets in X^*, as the unit ball is not compact in infinite dimensions. Applications include optimal control in function spaces, where the transform converts constrained problems into unconstrained ones.^[31] The Legendre transform finds significant use in partial differential equations, particularly Hamilton-Jacobi equations derived from action functionals. For a convex Hamiltonian H(x, p), the associated Lagrangian L(x, v) = \sup_p \{ \langle p, v \rangle - H(x, p) \} allows solutions to the Hamilton-Jacobi equation u_t + H(x, Du) = 0 to be represented as value functions u(x, t) = \inf_{\gamma} \int_0^t L(\gamma(s), \dot{\gamma}(s)) \, ds + u_0(\gamma(0)), where the infimum is over trajectories \gamma with \gamma(t) = x. This formulation, rooted in optimal control and viscosity solutions, characterizes the solution via the dynamic programming principle and extends to discounted problems \lambda u + H(x, Du) = 0, with u(x) = \inf \int_0^\infty e^{-\lambda s} L(\gamma(s), \dot{\gamma}(s)) \, ds. Such representations are crucial in homogenization and weak KAM theory on manifolds.^[32] Challenges arise in non-reflexive Banach spaces, where the canonical embedding X \hookrightarrow X^{**} fails, preventing automatic biconjugacy without additional assumptions. In such spaces, the double transform f^{**} may strictly contain f, complicating duality and requiring conditions like uniform convexity or Gateaux differentiability to ensure the Legendre map is well-behaved and single-valued. These issues limit applications in non-reflexive settings, such as certain L^p spaces for p \neq 2, unless reflexivity is imposed or alternative topologies are used.^[31]

Transformations under operations

The Legendre transform exhibits specific behaviors under various algebraic operations, preserving key structural properties of convex functions while adjusting the dual accordingly. These transformations are fundamental in convex analysis and facilitate the analysis of functions under changes in variables or compositions. Under translation, if a convex function f is shifted by a vector a, so that g(x) = f(x + a), then the Legendre transform of g is given by

g^*(p) = f^*(p) - p \cdot a.

This property follows from the definition of the conjugate via a change of variables in the supremum expression. For scaling operations with \lambda > 0, consider the rescaled function h(x) = \lambda f(x / \lambda). The Legendre transform simplifies to

h^*(p) = \lambda f^*(p).

This reflects the homogeneity adjustment, where the scaling factor \lambda directly multiplies the original conjugate, maintaining the overall degree of the function. Such scalings are particularly useful for analyzing positively homogeneous functions of degree one, whose conjugates are indicators of convex sets. Under linear transformations, suppose y = A x where A is an invertible linear map, and define the composed function k(y) = f(A^{-1} y). The Legendre transform of k relates to that of f via the adjoint operator A^* (the transpose for real matrices) as

k^*(p) = f^*(A^* p).

Equivalently, k^*(A^* p) = f^*((A^*)^2 p) does not generally simplify further unless A is orthogonal, but the form highlights the contravariant nature of the dual under linear changes. This property extends the transform to vector spaces with non-standard bases or metrics. The infimal convolution of two convex functions f and g, defined as (f \Box g)(x) = \inf_{u + v = x} \{f(u) + g(v)\}, has a Legendre transform that is the pointwise sum of the individual conjugates:

(f \Box g)^*(p) = f^*(p) + g^*(p).

This duality interchanges the infimal convolution in the primal domain with addition in the dual, a key result in optimization for decomposing problems into simpler subproblems. For certain reciprocal forms, if f(x) = 1 / g(1/x) where g is convex and positive, the Legendre transform exhibits an inverse relation under reciprocity, such that the dual of f connects back to a scaled or adjusted version of g^*, often appearing in contexts like thermodynamic potentials with inverse variables. This reciprocity underscores the transform's role in encoding complementary information symmetrically.

References

[1]
[PDF] Conjugates and Legendre transforms of convex functions
The one-to-oneness is equivalent to differentiability and strict convexity, plus a condition that the function become infinitely steep near boundary points of ...
[2]
[PDF] Legendre-Fenchel transforms in a nutshell - NC State ISE
All the results contained in this report can be found with much more mathematical details and rigor in R.T. Rockafellar, Convex Analysis, Princeton University ...
[3]
[PDF] The Legendre Transform 2 - Arizona Math
Jan 31, 2000 · The Transform. We shall define the Legendre transform. Let f : Rn !R be a convex function, i.e. r2f > 0 where rf is the hessian.
[4]
[PDF] Legendre Transforms - andrew.cmu.ed
Dec 26, 2012 · A Legendre transform is a procedure for expressing the information content of some function by using a different independent variable, namely, ...
[5]
[PDF] Making sense of the Legendre transform
The Legendre transform is a powerful tool in theoretical physics and plays an important role in classical mechanics, statistical mechanics, ...
[6]
[PDF] A Primer on the Legendre Transformation
Nov 21, 2023 · Section 2 discusses three properties of the Legendre transformation that are used subsequently and presents two more mathematical examples to ...
[7]
[PDF] A table of Legendre-transformation pairs with methodologies ... - arXiv
Aug 9, 2022 · The Legendre transformation, which relates curves in a pair of two-dimensional spaces, was introduced by Adrien-Marie Legendre (1752–1833). It ...Missing: formalized Laplace
[8]
None
### Summary of Relevant Sections from Field Theory Chapter 5
[9]
Convex Analysis - R. Tyrrell Rockafellar - Google Books
Available for the first time in paperback, R. Tyrrell Rockafellar's classic study presents readers with a coherent branch of nonlinear mathematical analysis ...
[10]
[PDF] LEGENDRE TRANSFORM AND APPLICATIONS TO ... - Hal-Inria
Indeed, g is essentially smooth if and only if g∗ is strictly convex on every convex subset of dom ∂g; see [31, Theorem 26.3]. The last property is known ...Missing: invertibility | Show results with:invertibility
[11]
None
Below is a merged summary of the Legendre Transform and Convex Conjugate concepts from "Convex Optimization" by Boyd & Vandenberghe, based on the provided segments. To retain all information in a dense and organized manner, I will use a table format in CSV style, followed by a concise narrative summary. The table captures key details such as section references, definitions, geometric interpretations, envelopes, supporting hyperplanes, key quotes, and useful URLs for each relevant segment.
[12]
[PDF] Math 115 (2006-2007) Yum-Tong Siu 1 Legendre Transformation ...
For the second method, the family of geometric objects is a family of straight lines and higher-order contact means that a line in the family is tangential ...
[13]
[PDF] The concept of duality in convex analysis, and the characterization ...
is affine linear, and therefore so is G2, so we write that G2(a, A) = B2(a, A) + V2 for B2 ∈ GLn+1,V2 ∈ Rn+1. Page 9. DUALITY IN CONVEX ANALYSIS. 669. Moreover, ...
[14]
https://legacy-www.math.harvard.edu/archive/115_fall_06/legendre_transformation.pdf
[15]
Legendre Transformation and Thermodynamic Potentials | Resonance
May 7, 2024 · Not only common potentials—internal energy, Helmholtz free energy, Gibbs potential ... H B Callen, Thermodynamics and an Introduction to ...
[16]
[PDF] 8.044 Lecture Notes Chapter 5: Thermodynamics, Part 2
So to go back and forth between E and H we Legendre transform, either way. Maxwell Relations. dE = T dS − P dV. dH = T dS + V dP. =⇒.
[17]
[PDF] Thermodynamic Potentials and Maxwell's Relations
Feb 25, 2003 · In a Legendre transform, to replace one independent variable with its conjugate, a new function ζ is defined by the addition or subtraction of ...
[18]
[PDF] 4. The Hamiltonian Formalism - DAMTP
There is a mathematical trick to do this, known as the Legendre transform. To describe this, consider an arbitrary function f(x, y) so that the total derivative ...
[19]
[PDF] Legendre transforms
Feb 15, 2019 · The. Legendre transformation connects two ways of specifying the same physics, via functions of two related (“conjugate”) variables. Table 1 ...<|control11|><|separator|>
[20]
10. Canonical Transformations - Galileo and Einstein
... variables are said to be canonically conjugate. The Poisson bracket is invariant under a canonical transformation, meaning. [f,g]p,q=[f,g]P,Q. Let's begin by ...
[21]
[PDF] Part I - Duality of Production, Cost, and Profit Functions
The theory establishing the dual relation between cost functions and production functions was introduced into economics by Shephard. (1953), who drew heavily ...
[22]
(PDF) Economic Analysis of Technology and Properties of Legendre ...
Aug 6, 2025 · ... Duality Theorem of Legendre-Fenchel transformations ... The pseudo production and pseudo cost functions are identified as dual functions of ...
[23]
[PDF] DUAL APPROACHES TO UTILITY
6 This procedure is known as the Legendre transform. 10. Page 10. ensure that the minimum in (1) is attained for every p. (U1) u(.) is continuous. Given this ...<|separator|>
[24]
[PDF] The Legendre Transformation in Modern Optimization
Abstract The Legendre transformation (LET) is a product of a general du- ality principle: any smooth curve is, on the one hand, a locus of pairs, which.
[25]
Large Deviations Techniques and Applications - SpringerLink
Amir Dembo and Ofer Zeitouni, two of the leading researchers in the field ... Available as PDF; Read on any device; Instant download; Own it forever. Buy ...
[26]
Variational Characterization of Free Energy: Theory and Algorithms
A well-known thermodynamic principle states that the free energy is the Legendre transform of the entropy. The following variant of this principle is due to ...
[27]
Connecting Information Geometry and Geometric Mechanics - MDPI
Convex duality and the Legendre transform play a key role in both ... A Riemannian manifold M with metric g i j admits biorthogonal coordinates if and ...
[28]
[PDF] Tonelli Lagrangians and Hamiltonians on Compact Manifolds
Mar 2, 2015 · Let us describe how to define this new system and what its relation is with the. Lagrangian one. A standard tool in the study of convex ...
[29]
ISOMETRIES FOR THE LEGENDRE-FENCHEL TRANSFORM
obtain isometries for the infinite dimensional case (Hilbert space and reflexive. Banach space), but this time they correspond to topologies finer than the ...
[30]
[PDF] HAMILTON–JACOBI EQUATIONS: THEORY AND APPLICATIONS
... Legendre's transform ... action functionals . . . . . . . . . . 243. 3. Characterization of the Legendre transform ...