The empirical kernel covariance operator is self-adjoint and positive semidefinite. Maximize subject to . The first variation of the Lagrange multiplier functional gives
so . Taking the inner product with gives
In the long-wavelength description of an antiferromagnet, is the antiferromagnetic spin wave velocity and controls the stiffness or strength of quantum fluctuations. The field is an auxiliary Lagrange multiplier enforcing the fixed-length constraint on the order parameter. At the translation-invariant saddle, shifts every propagator denominator and is the excitation gap, which explains the name gap equation.
The Large-N expansion makes fluctuations of the auxiliary field relatively small. Its leading saddle-point condition is then self-consistent and becomes the displayed gap equation.
At a regular point of the optimization constraint, meaning , the necessary Lagrange multiplier condition is
Indeed, every allowed tangent vector to the level set satisfies . At a constrained extremum, the directional derivative must vanish too. Thus both gradients are normal to the same tangent line and are parallel. This includes . At a singular point with , the displayed condition need not be necessary; those points require a separate examination.
For , the gradient never vanishes on the circle. The Lagrange multiplier equations are
If , the two possibilities are . If , then , and the first equation gives . Combining this with yields
Put
Both and are positive, giving the other four points and . To classify all six, restrict to the circle by eliminating :
Its derivatives are and . In the interior, and . At either endpoint ; along the circle, has a strict local minimum at the left endpoint and a strict local maximum at the right endpoint. Consequently the six constrained extrema are
The circle is a compact set, so the extreme value theorem guarantees that its global minimum and global maximum occur among these six points. Their values give the corrected sharp bound
The printed lower bound is false. A particularly short counterexample is
The sole local minimum in part (a), , has value , larger than values on this circle; it is therefore not a global minimum. In fact , so has no global minimum on the plane.
Figure 1.
All six extrema on the constraint circle and the corrected minimum of the restricted polynomial
. Details.
Let and . The unit sphere is compact, so the continuous function has a global maximum and global minimum. Since on the unit sphere, all such points satisfy the Lagrange multiplier equations
Thus and either or .
If , then , giving . If , then and , giving . These exhaust the candidates, so
Assume are continuously differentiable near the point and . The implicit function theorem supplies a local function with
At an interior extremum of , the multivariable chain rule gives
Multiply these identities by and , respectively, and subtract. The terms containing cancel, giving
No assumption is needed for this elimination; dividing by would unnecessarily exclude valid cases. If , this particular local graph argument is unavailable and a different chart or the full Lagrange multiplier condition is needed.
Use a Lagrange multiplier to maximize subject to , with . The equations are
The second equation gives . Substituting into the first gives , hence . The constraint then gives , so
This is the same aspect ratio as the minimum-area closed cylinder at fixed volume. To confirm maximality directly, eliminate : on . Its only stationary point is the value above, its second derivative there is negative, and the volume tends to zero at either end of the interval.
Maximize the total surface area with the constraint . The Lagrange multiplier equations are
From the second, . Substitute into the first and multiply by to obtain . Thus the aspect ratio satisfies . Since ,
This is the maximum-area cylinder inscribed in a sphere. The dimensions and area, although not needed to specify the ratio, are
There is only one positive stationary ratio. At the degenerate boundaries and , the limiting areas are zero and , both below this value, confirming the maximum.
For two independent smooth constraints , the tangent direction is orthogonal to both gradients. At a constrained stationary point, the gradient of the objective is also orthogonal to that direction, hence lies in their span: . This is the two-constraint Lagrange multiplier condition. Alternatively, distance extrema on a central plane section of an ellipsoid follow from a two-dimensional eigenvalue problem.
Minimizing distance is equivalent to minimizing . Put and . The Lagrange multiplier equations are
Adding the latter two gives . If , those equations give ; the constraints then give and . Otherwise and . The ellipsoidal constraint gives , and hence
The intersection is compact. The two constraint gradients are independent there: if , then , and forces , contradicting . Thus every global extremum is among these candidates. Taking square roots gives
Expand . The Lagrange multiplier equations for are and , together with the constraint . The first two imply and . The value would make the latter equation inconsistent, so division is allowed:
Insert these into and multiply by . Simplification gives , hence
The gradient of the constraint is never zero, so the Lagrange multiplier condition is necessary at every constrained extremum. To establish a global minimum, parametrize the whole parabola by . The restricted function becomes
It is coercive: as , so it attains a global minimum. Moreover , making it strictly convex and ensuring that this minimum is unique. This proves more than the permitted assumption of one real multiplier root.
The level sets of are nested ellipses centered at the origin. In the rotated coordinates , , they have equation , with the longer axis along . The smallest one meeting the parabola is tangent to it at the unique minimizing point, approximately , where and .
Figure 1.
The smallest quadratic contour touching a parabolic constraint
. Original level-set plot. The highlighted ellipse is tangent to the parabola at the unique global minimum; larger quadratic contours intersect the constraint. Details.
The base costs , the glass front costs , the wooden rear costs , and the other two walls cost . Thus minimize subject to . With the Lagrange multiplier convention , the equations are
Multiply these by , respectively. Comparing their equal right sides gives . Since all dimensions are positive, and . The volume condition then yields
This stationary shape is the unique global minimum, not merely a candidate: the arithmetic-geometric mean inequality applied to the three cost terms gives
with equality exactly at those ratios. Therefore the least cost, in pounds in the specified units, is
This is a weighted open-box minimization, with each pair of opposite walls combined into its appropriate positive face weight.
The Lagrange multiplier equations are , , together with . If , then and ; if , then and . Thus
The constraint is compact and its gradient never vanishes, so these exhaust the possible global extrema. Equivalently, substituting , gives along the boundary.
Minimize the squared distance subject to . The Lagrange multiplier condition gives and . Enforcing the line constraint gives , hence
The displacement to this point is parallel to the line's normal vector , so it is the orthogonal projection. The squared distance restricted to the line has positive quadratic coefficient , proving this stationary point is the unique global minimum.
Let denote the stated cost per area of the cylindrical wall. Charge for the curved wall and cap as specified; no floor cost is supplied. The volume and cost are and . To minimize cost at fixed volume, apply a Lagrange multiplier to . Differentiation with respect to gives . The derivative then gives , reducing to . Thus the optimal cylinder and hemispherical cap with unequal surface costs has
For a global check, eliminate to obtain . Its second derivative is positive for , and its critical radius gives , so the constrained optimum lies inside the physically feasible range and is the unique minimum.
For the Lagrange multiplier equations, use . They give , , and . If , then , implying and therefore , contradicting . Thus only , remain.
The circle is compact and the function is continuous, so global extrema exist. On it, eliminating gives , which is increasing for since its derivative is . Therefore
These constrained extrema are flat to quadratic order along the circle, so a vanishing second derivative alone would not classify them.
Minimize subject to , , where and . The Lagrange multiplier equation gives . Since , , and , the constraints imply , . Hence
No explicit parametrization of the line was needed. For any other feasible point, is orthogonal to both normals and therefore to . The Pythagorean theorem gives , proving the unique global minimum. This is the geometric principle behind the minimum-norm least-squares solution.
Set . The gradient of never vanishes, so the Lagrange multiplier equations apply at a constrained stationary point. They are , , and . Eliminating gives
With , this becomes . The quadratic factor is , so the only real possibility is . Therefore
The level curve through this point is the ellipse , meeting the coordinate axes at and . Its tangent line at has slope , equal to the derivative of the constraint parabola there. The two curves are tangent.
Figure 1.
The smallest elliptical contour meeting the constraint is tangent at the minimum
. The highlighted contour is and the constraint is . Its four axis intercepts and the unique contact point are marked. Details.
The parabola touches this ellipse from outside; every other point on the constraint lies on a larger contour. An exact check of this geometric argument is
with equality only at . Thus the point is the unique constrained global minimum.
Fix the total area at . Apply Lagrange multipliers to with constraint . The two derivative equations are
Since , the second gives . Substitution in the first gives . The constraint then yields . Therefore
For completeness, eliminating makes on . Its derivative changes from positive to negative at the calculated radius and its volume tends to zero at both domain endpoints. Thus this stationary cylinder is indeed the maximum. It is the fixed-area case of optimal aspect ratios of closed cylinders.
Divide the objective area by , which does not change its maximizer, and use with constraint . The Lagrange multiplier equations are
At a nondegenerate cylinder , the second gives . Eliminating it gives . Hence satisfies . Select its positive root:
This is the sphere-inscribed case of optimal aspect ratios of closed cylinders. One may check maximality from : its derivative has the sign of , so it changes from positive to negative at the selected root.
Take positive dimensions measured in metres. The volume is and the external surface area constraint is . There is one base and four walls, with no lid. The Lagrange multiplier equations for are
The first equation implies . Multiply the first by and the second by and subtract; this gives , hence . The last equation gives . Substituting into the first yields and therefore .
The constraint now reads . Thus
For a global check, apply the arithmetic-geometric mean inequality to , and , whose sum is . Their product is , so no feasible box has . Equality holds exactly at the dimensions found above. This is the maximum-volume open rectangular box at fixed surface area.