Write and impose tracelessness by a Lagrange multiplier . Differentiating givesFor , each eigenvalue is a real root of the same nonzero cubic polynomial, so at most three distinct values occur at any stationary point, in particular at any existing minimum. If the multiplicities are , the unbroken subgroup is the block centralizerIts connected Lie algebra is . The partitions allowed by this upper bound are , , , , and . These are candidate patterns, not a proof that every one is a stable minimum for given couplings. For three distinct roots, the absent quadratic coefficient also implies that their unweighted sum is zero; tracelessness imposes an additional multiplicity-weighted relation.
The unrestricted printed claim has an exception. If , and , every traceless matrix with is a global minimum. For examplehas five distinct eigenvalues and attains that minimum. Its unbroken subgroup is the maximal torus . The cubic argument therefore requires a nonzero , or another condition excluding the identically vanishing stationarity polynomial. Couplings for which the potential is unbounded below have no global minimum to classify.
The constraint implies , so and its gradient never vanishes on the feasible set. The Lagrange multiplier equations giveThus , , , and feasibility gives . The two stationary points and their values areThey are both global minima. Indeed, on the constraint, completing the square givesEquality holds only at the two displayed points. There are no other constrained stationary points, and the objective is unbounded above as .
For positive edge lengths the volume is and the surface area constraint is . Introduce a Lagrange multiplier :The stationary point equations areMultiplying these by , respectively, and equating their left sides gives and . Because all edge lengths are positive, the equations imply ; hence . The constraint gives .
To establish a global maximum, rather than only a stationary point, apply the arithmetic-geometric mean inequality to :Equality holds exactly when , or . Therefore
Let and . Subtracting the two constraint equations gives , hence and . The feasible set is therefore a circle.
For Lagrange multipliers , the stationarity equation becomesAt , the first equation gives . The second ensures and , so the third gives . The constraints then force . Both points do satisfy the multiplier equations, taking .
There are no missing singular constraint points: on the feasible circle, the two constraint gradients and are linearly independent, since . Thus the complete list of constrained critical points and critical values isThe bound on the feasible circle shows that the positive value is the global maximum and the negative value is the global minimum.
On the unit circle the constraint gradient never vanishes. The Lagrange multiplier equations areA nonzero solution of the first two equations requiresso or . Multiplying the first equation by , the second by , and adding shows . Thus each candidate value of equals its multiplier.
For , , giving . For , , giving . Since a continuous function attains its extrema on the compact set given by the unit circle and these exhaust the Lagrange multiplier candidates,Each changes both coordinates together.
The constraint is a compact set, so the continuous objective attains both extrema by the extreme value theorem. Use Lagrange multipliers with :Since the second component makes , the two equations give , or . Substitution into the constraint yieldswith matching signs. The objective values are thereforeThe constraint gradient never vanishes on the ellipse, so these are all the possible extrema. As a direct check, set , ; then and the Cauchy-Schwarz inequality gives , with equality at the same points.
The continuous function attains a global maximum and minimum on the closed disc by the extreme value theorem. All four unconstrained stationary points have squared distance eight, so none lies in its interior. The extrema must therefore lie on .
The constraint gradient is nonzero on that circle, so Lagrange multipliers apply:Subtracting gives . If , the constraint gives . Otherwise ; substitution into either equation gives . This is impossible, since on the circle. Thus the diagonal points are the only candidates. Comparing their values gives
Use the LagrangianThe Lagrange multiplier equations areFeasibility makes , so . The second equation becomes , forcing and then . The constraint gradient is nonzero on the feasible set, so the multiplier calculation has no missing singular feasible point.
To verify the global minimum rather than just stationarity, eliminate :Equality holds exactly at or , with minimum value .
Separate inactive and active optimization constraints. At an interior feasible point with , a differentiable local minimum must satisfy . Retain only solutions lying in the interior.
On the boundary , if , apply the Lagrange multiplier condition. Using the optimization Lagrangian , the Karush-Kuhn-Tucker conditions areThe sign follows from feasible motion into : the directional derivative along at a minimum cannot be negative. Thus a boundary candidate has and a nonnegative Lagrange multiplier. Boundary points with must be examined separately, because the usual constraint qualification fails there.
Solve the interior and boundary systems, check membership in the feasible set, and establish which candidates are minima using second-order tests or a direct global comparison. The Karush-Kuhn-Tucker conditions alone are not sufficient for general nonconvex functions or feasible sets. On an unbounded feasible set, also check whether an infimum is attained and whether escaping sequences can improve the objective.
The objective is continuous on the compact disk, so a global maximum exists. Its gradient is , which is never zero, so a maximum cannot lie in the interior. It must have .
Both coordinates of a maximizing point are positive. Reflecting a negative coordinate to its positive counterpart strictly increases the objective without changing the constraint; at an endpoint with one coordinate zero, a small move along the circle into the positive quadrant also increases the objective to first order. Thus it suffices to consider .
The boundary Lagrange multiplier equations are and . Eliminate to obtain the two-dimensional nonlinear system of equationsKeeping this quadrant restriction excludes the boundary minimum that an unconstrained elimination could also produce. Under the stated uniqueness assumption for the admissible solution, the global maximum must be that solution. Approximate it with multivariable Newton iteration, solving and setting , with a suitable safeguard if needed. For example its explicit Jacobian matrix isNo numerical solution of the system is required.
Assume and are continuously differentiable near the point. At a regular feasible point, meaning and , a differentiable constrained local extremum satisfies the Lagrange multiplier equationsThey arise by differentiating the optimization Lagrangian with respect to all coordinates and the multiplier. Geometrically, the objective has zero derivative along the tangent space of the constraint surface, so its gradient must lie in the normal direction.
Solve these equations for candidates, examine feasible points where the regularity condition fails separately, and compare objective values or use a second-order test restricted to feasible directions. The equations locate stationary points; they do not by themselves establish which are minima or whether a global minimum is attained.
Use and . Every feasible point has , so cannot vanish. The Lagrange multiplier equations areThe first gives . If , the second would give , a contradiction. Therefore , and the constraint gives . The only constrained stationary points are .
Here the constraint surface is compact and regular, so a global minimum exists and satisfies the Lagrange multiplier equationsKeep the zero-coordinate cases. If , the two points are , with value . If , the two points are , with value . If both coordinates are nonzero, the equations give . The constraint becomes , whose only positive solution is . These are the four points , all with value .
These cases exhaust the stated eight constrained stationary points. Comparing their values givesFor additional classification, substituting gives , withIt decreases from to and increases from to . Thus the two -axis points are constrained local maxima, the two -axis points are global maxima, and the four mixed-coordinate points are global minima.
Both terms and are nonnegative, and they vanish simultaneously only at . That point satisfies . Consequentlyand the global minimum is unique. No Lagrange multiplier calculation is needed: nonnegativity gives a lower bound attained at a feasible point.
The origin is feasible and has objective zero, while a Euclidean norm is nonnegative. Thus the minimum is , and the complete argmin isIf is injective, this set contains only the origin. If it has a nontrivial kernel, every feasible kernel vector is also a minimizer. No Lagrange multipliers are needed.
Interpret the norms as Euclidean norms and set . This is a real symmetric matrix with , hence a positive semidefinite matrix. Squaring the nonnegative objective does not change its argmin. Let be the smallest eigenvalue of and its eigenspace.
First minimize on the unit sphere. The Lagrange multiplier function has stationarity condition . Multiplication by shows that the squared objective at such a point is . The spectral theorem for real symmetric matrices also gives, for every unit vector, , with equality exactly in . This proves global minimality, rather than merely finding stationary candidates.
If , every point with has a strictly smaller objective after rescaling to unit length. ThereforeIf , the kernel of a linear map satisfies , since implies , and the converse is immediate. The minimum is then zero, attained precisely atIn this case minimizers need not have unit length. This distinction is essential for a rank-deficient , including any real matrix with . The result is the minimum image norm outside a Euclidean unit ball, governed by the least singular value.
Phase-space action Created 2026-10-05 Updated 2026-10-06
A phase-space action treats canonical variables as independent fields. Varying them gives Hamilton's equations; Lagrange multipliers can impose constraints.
A nonconvex primal problem can have a positive duality gap even when both optima are attained. Choose the finite set , the objective function values , , , and the equality optimization constraint . The only feasible point is , so .
With unrestricted equality Lagrange multiplier , the three possible optimization Lagrangian values and their minimum areThe concave Lagrange dual function has unique maximizer , with . Thus the duality gap is . At the optimal Lagrange multiplier, both inner minimizers and violate the equality; the dual optimum does not recover a primal optimizer.
Why can no slope close the gap? Any affine function lying below the two endpoint costs obeys and . Adding gives , so its value at residual zero cannot reach the feasible cost . A convex combination of the endpoint cost-residual pairs with equal weights has residual zero and average cost zero. This averaged pair is unattainable: the actual decision costs . This geometry explains precisely what the dual bounds fail to distinguish. Nonconvexity does not force a positive duality gap: changing only to makes both optimal values zero on the same nonconvex domain.
Let be a symmetric positive-definite matrix, , and . Thenis the unique solution of minimizing the quadratic form under the affine optimization constraint . The minimum value is .
The denominator is positive because is a positive-definite matrix. The Lagrange multiplier equations give the displayed candidate and verify the constraint. For a global proof, any feasible perturbation satisfies . HenceThis is positive for every nonzero . The result also handles several independent quadratic blocks by taking to be their block diagonal matrix and concatenating their variables.
Primal-dual interior-point method 2026-10-31
A primal-dual interior-point method solves perturbed KKT conditions while updating both the primal variables and Lagrange multipliers. For inequalities written as , a logarithmic-barrier path has positive and perturbed complementarity . A Newton system for these coupled equations gives primal and dual corrections together, with damping to preserve positivity. This can avoid explicitly forming the most ill-conditioned eliminated barrier Hessian, although scaling and regularity still matter.
