How To Find Critical Numbers Subject To

How to Find Critical Numbers Subject to Constraints: A Step‑by‑Step Guide

Finding critical numbers is a fundamental skill in calculus that helps you locate where a function’s derivative is zero or undefined—points that often correspond to local maxima, minima, or saddle points. This leads to when you add a constraint (for example, optimizing a function while staying on a curve or surface), the process shifts slightly but remains grounded in the same core ideas. This article walks you through the theory, the procedural steps, and concrete examples so you can confidently determine critical numbers both for unrestricted functions and for problems subject to constraints Simple as that..

What Are Critical Numbers?

A critical number (or critical point) of a function (f(x)) is any value (c) in the domain of (f) where either:

1. (f'(c) = 0) (the derivative equals zero), or
2. (f'(c)) does not exist (the derivative is undefined).

These points are important because, according to Fermat’s Theorem, if (f) has a local extremum at (c) and (f) is differentiable at (c), then (c) must satisfy (f'(c)=0). Points where the derivative fails to exist can also host extrema or cusps, so they must be examined as well The details matter here..

When a problem includes a constraint—say, you want to extremize (f(x,y)) subject to (g(x,y)=0)—the critical numbers are found by solving a system that incorporates both the objective function and the constraint, most commonly via the Lagrange multiplier method.

Steps to Find Critical Numbers for an Unrestricted Function

For a single‑variable function (f(x)), follow this checklist:

1. Determine the domain of (f). Exclude any points where the function itself is undefined (e.g., division by zero, log of non‑positive numbers).
2. Compute the derivative (f'(x)) using appropriate rules (power, product, quotient, chain, etc.).
3. Set the derivative equal to zero and solve for (x). These solutions are candidate critical numbers.
4. Identify where the derivative is undefined (but the original function is defined). Add those (x)-values to the list.
5. Verify each candidate lies within the domain of (f). Discard any that fall outside.
6. (Optional) Classify each critical number using the first or second derivative test to determine if it corresponds to a maximum, minimum, or neither.

Example 1: Polynomial Function

Find the critical numbers of (f(x)=x^{3}-3x^{2}+2).

1. Domain: all real numbers (\mathbb{R}).
2. Derivative: (f'(x)=3x^{2}-6x).
3. Set to zero: (3x^{2}-6x=0 \Rightarrow 3x(x-2)=0 \Rightarrow x=0) or (x=2).
4. Derivative is a polynomial, so it is defined everywhere.
5. Both (0) and (2) belong to (\mathbb{R}).

Critical numbers: (x=0,; x=2).

Example 2: Rational Function with a Hole

Find the critical numbers of (f(x)=\dfrac{x^{2}}{x-1}).

1. Domain: (x\neq 1) (denominator zero).
2. Derivative via quotient rule:
   [ f'(x)=\frac{(2x)(x-1)-x^{2}(1)}{(x-1)^{2}}=\frac{2x^{2}-2x-x^{2}}{(x-1)^{2}}=\frac{x^{2}-2x}{(x-1)^{2}}=\frac{x(x-2)}{(x-1)^{2}}. ]
3. Set numerator to zero: (x(x-2)=0 \Rightarrow x=0) or (x=2).
4. Derivative undefined where denominator zero: (x=1). Even so, (x=1) is not in the domain of (f), so it is not a critical number.
5. Both (0) and (2) are in the domain.

Critical numbers: (x=0,; x=2).

Critical Numbers Subject to Constraints: Lagrange Multipliers

When you must optimize a function (f(x,y,\dots)) subject to an equality constraint (g(x,y,\dots)=0), the critical numbers of the Lagrangian

[ \mathcal{L}(x,y,\dots,\lambda)=f(x,y,\dots)-\lambda,g(x,y,\dots) ]

give the candidates. The method relies on the geometric idea that at an extremum, the gradient of (f) is parallel to the gradient of (g).

Step‑by‑Step Lagrange Multiplier Procedure

1. Write the Lagrangian: (\mathcal{L}=f-\lambda g).
2. Compute partial derivatives with respect to each variable and (\lambda): [ \frac{\partial \mathcal{L}}{\partial x}=0,\quad \frac{\partial \mathcal{L}}{\partial y}=0,\quad \dots,\quad \frac{\partial \mathcal{L}}{\partial \lambda}=0. ] The last equation simply returns the constraint (g=0).
3. Solve the resulting system of equations for the variables and (\lambda).
4. Check each solution against the domain of the original problem (e.g., avoid division by zero, ensure variables stay within allowed intervals).
5. Evaluate (f) at each feasible solution to determine which yields a maximum, minimum, or saddle point (if needed, use bordered Hessian or second‑derivative test for constrained problems).

Example 3: Constrained Optimization with Two Variables

Find the extreme values of (f(x,y)=x^{2}+y^{2}) subject to the constraint (g(x,y)=x+y-1=0) (i.In real terms, e. , the line (x+y=1)) Practical, not theoretical..

1. Lagrangian: (\mathcal{L}(x,y,\lambda)=x^{2}+y^{2}-\lambda(x+y-1)).
2. Partial derivatives: [ \frac{\partial \mathcal{L}}{\partial x}=2x-\lambda=0 \quad\Rightarrow\quad \lambda=2x, ] [ \frac{\partial \mathcal{L}}{\partial y}=2y-\lambda=0 \quad\Rightarrow\quad \lambda=2y, ] [ \frac{\partial \mathcal{L}}{\partial \lambda}=-(x+y-1)=0 \quad\Rightarrow\quad x+y=1. ]
3. From the first two equations, (2x=2y) ⇒ (x=y).
4. Sub

stitute (x=y) into the constraint (x+y=1): [ x+x=1 ;\Rightarrow; 2x=1 ;\Rightarrow; x=\frac{1}{2},; y=\frac{1}{2}. ] 5. So naturally, the corresponding multiplier is (\lambda=2x=1). Here's the thing — 6. Practically speaking, evaluate (f) at the candidate: [ f! \left(\tfrac{1}{2},\tfrac{1}{2}\right)=\left(\tfrac{1}{2}\right)^{2}+\left(\tfrac{1}{2}\right)^{2}=\frac{1}{2}. ] Because (f(x,y)=x^{2}+y^{2}) represents the squared distance from the origin and the constraint is an unbounded line, this single critical point gives the absolute minimum (\frac{1}{2}). There is no maximum on the line (x+y=1) since (f) grows without bound as (|x|,|y|\to\infty).

Example 4: Constrained Optimization with a Non‑Linear Constraint

Find the extreme values of (f(x,y)=xy) subject to (g(x,y)=x^{2}+y^{2}-1=0) (the unit circle).

1. Lagrangian: (\mathcal{L}(x,y,\lambda)=xy-\lambda(x^{2}+y^{2}-1)).
2. Partial derivatives: [ \frac{\partial\mathcal{L}}{\partial x}=y-2\lambda x=0,\qquad \frac{\partial\mathcal{L}}{\partial y}=x-2\lambda y=0,\qquad \frac{\partial\mathcal{L}}{\partial \lambda}=-(x^{2}+y^{2}-1)=0. ]
3. Solve the system:
   From the first two equations, (y=2\lambda x) and (x=2\lambda y). Multiplying them gives (xy=4\lambda^{2}xy).
   • If (xy\neq0), then (4\lambda^{2}=1\Rightarrow\lambda=\pm\frac{1}{2}).
     • (\lambda=\frac{1}{2}): (y=x). Constraint (\Rightarrow 2x^{2}=1\Rightarrow x=y=\pm\frac{1}{\sqrt{2}}).
     • (\lambda=-\frac{1}{2}): (y=-x). Constraint (\Rightarrow 2x^{2}=1\Rightarrow x=\pm\frac{1}{\sqrt{2}},; y=\mp\frac{1}{\sqrt{2}}).
   • If (xy=0), then either (x=0) or (y=0). Constraint gives the points ((0,\pm1)) and ((\pm1,0)) (these correspond to (\lambda=0)).
4. Evaluate (f): [ \begin{array}{c|c} (x,y) & f(x,y)=xy \ \hline \left(\frac{1}{\sqrt{2}},\frac{1}{\sqrt{2}}\right),; \left(-\frac{1}{\sqrt{2}},-\frac{1}{\sqrt{2}}\right) & \frac{1}{2} \quad\text{(maximum)}\[4pt] \left(\frac{1}{\sqrt{2}},-\frac{1}{\sqrt{2}}\right),; \left(-\frac{1}{\sqrt{2}},\frac{1}{\sqrt{2}}\right) & -\frac{1}{2} \quad\text{(minimum)}\[4pt] (\pm1,0),;(0,\pm1) & 0 \quad\text{(saddle points on the constraint)} \end{array} ]

Beyond Equality Constraints: KKT Conditions

Real‑world problems often involve inequality constraints (g_i(x)\le 0) (e.g., budget limits, non‑negativity) And that's really what it comes down to..

1. Stationarity: (\nabla f(\mathbf{x}^) + \sum \mu_i \nabla g_i(\mathbf{x}^) = 0).
2. Primal feasibility: (g_i(\mathbf{x}^*) \le 0).
3. Dual feasibility: (\mu_i \ge 0).
4. Complementary slackness: (\mu_i g_i(\mathbf{x}^*) = 0) for all (i).

A constraint is active (binding) if (g_i=0) and its multiplier (\mu_i>0); it is inactive if (g_i<0) and (\mu_i=0). This framework reduces to Lagrange multipliers when all constraints are equalities (where (\mu_i) are unrestricted in sign).

Practical Checklist for Finding Critical Numbers

| Multi-variable with equality constraints (h(x,y,\dots)=0) | Use Lagrange multipliers: solve (\nabla f = \lambda \nabla h) together with (h=0). And | | Multi-variable with inequality constraints (g_i(x,y,\dots) \le 0) | Apply KKT conditions: stationarity, primal/dual feasibility, and complementary slackness. But | | Boundary or corner points | Always check boundary points and corners separately—extrema can occur there even if (\nabla f \neq 0). | | Domain restrictions | Verify that critical points lie within the allowed domain; discard any that violate constraints Most people skip this — try not to..

Conclusion

Finding critical numbers is a systematic process that adapts to the structure of the problem at hand. By combining analytical techniques like Lagrange multipliers and KKT conditions with careful evaluation of boundaries and domain restrictions, we can confidently locate maxima, minima, and saddle points. Whether dealing with single-variable functions, multivariable expressions, or constrained optimization scenarios, the core principle remains the same: identify points where the function’s behavior changes—where derivatives vanish or fail to exist. This structured approach not only solves textbook exercises but also provides a solid foundation for tackling real-world optimization challenges in economics, engineering, and beyond.