Function Model Construction and Application

SECTION 1

Historical Context & Motivation

The practice of constructing mathematical models to describe observable phenomena is as old as mathematics itself. Ancient Babylonian scribes used quadratic relationships to compute areas and forecast agricultural yields, while Greek geometers implicitly worked with polynomial relationships when studying conic sections. The modern notion of a function as a rule that assigns each input exactly one output crystallized only over centuries of refinement, beginning with Leibniz's use of the Latin word functio in 1694 and culminating in Dirichlet's rigorous definition in the nineteenth century. This evolution transformed mathematics from a collection of geometric tricks into a powerful language for modeling the natural world.

c. 1800 BCE

Babylonian Quadratics

Babylonian mathematicians solve quadratic equations on clay tablets to model areas and commercial transactions, establishing the earliest polynomial relationships.

1637

Descartes' Coordinate Geometry

René Descartes publishes La Géométrie, uniting algebra and geometry through the coordinate plane and enabling polynomial curves to be expressed as equations.

1694

Leibniz Coins 'Function'

Gottfried Wilhelm Leibniz introduces the term function to describe quantities that depend on a variable, formalizing a concept implicit in calculus.

1795

Lagrange Interpolation

Joseph-Louis Lagrange publishes a systematic method for constructing a polynomial of minimal degree that passes through a given set of data points.

20th Century

Modern Modeling

Polynomial and rational function models become standard tools across engineering, economics, and the sciences, supported by computing technology that automates regression and curve fitting.

The central question that drives this lesson is deceptively simple: given a real-world scenario—perhaps described by a table of values, a verbal description, or a set of algebraic constraints—how do we construct an appropriate polynomial or rational function model, and how do we then use that model to make predictions and draw conclusions? The AP Precalculus framework places this skill at the heart of the polynomial and rational functions unit, because the ability to move fluently between representations—verbal, numerical, algebraic, and graphical—is the hallmark of mathematical maturity.

SECTION 2

Core Principles of Function Model Construction

Constructing a function model is not simply curve-fitting; it requires a deliberate, principled process in which you identify the type of behavior exhibited by the data or situation, select an appropriate function family, determine the specific parameters, and validate the model against known constraints. The following foundational ideas underpin every model-construction problem you will encounter on the AP Precalculus exam.

Function Family Selection

Determine whether the situation calls for a polynomial (finite turning points, continuous domain) or a rational function (asymptotic behavior, restricted domain). The end behavior, number of zeros, and presence of discontinuities guide this choice.

Degree and Structure

For polynomials, the degree determines the maximum number of turning points (n − 1) and zeros (n). For rational functions, the degrees of numerator and denominator control horizontal and slant asymptotes. Always choose the lowest degree that fits the data.

Zeros, Intercepts, and Multiplicity

Real zeros of the function correspond to x-intercepts. The multiplicity of each zero determines the graph's local behavior—whether it crosses or touches the axis. Zeros of the denominator in a rational model create vertical asymptotes or holes.

Leading Coefficient and Scaling

After determining zeros and structure, a leading coefficient (often denoted a) scales the model vertically. An additional known point—such as the y-intercept or another data value—is used to solve for a.

Validation and Interpretation

A constructed model must be checked against all given information: sign behavior, domain restrictions, and real-world reasonableness. Interpreting outputs in context—units, feasibility, limiting behavior—is the final and most critical step.

✦ KEY TAKEAWAY

Think of function model construction like designing a custom key. The function family is the blank key shape, the zeros and asymptotes are the specific notches you carve, and the leading coefficient is the final filing that ensures the key turns the lock. Each piece of information narrows the possibilities until exactly one model fits.

SECTION 3

Visual Explanation — From Zeros to Function

The diagram below illustrates the step-by-step process of constructing a cubic polynomial from a set of zeros and a known point. Observe how each piece of information progressively constrains the model until a unique function emerges.

The three colored dots on the x-axis represent the zeros x = −2, x = 1, and x = 4, which define the factored form of the cubic. The amber dot at (0, −8) is the additional known point used to solve for the leading coefficient a = −1, yielding f(x) = −(x + 2)(x − 1)(x − 4).

Notice how the cubic curve crosses the x-axis at each zero with multiplicity one, producing a sign change at every intercept. The end behavior—rising to the left and falling to the right—is consistent with a negative leading coefficient on an odd-degree polynomial. This visual check is an essential part of model validation: the graph must be consistent with every algebraic feature you built into the model.

SECTION 4

Mathematical Framework

The algebraic machinery for constructing polynomial and rational models rests on factored forms, because factored forms directly encode the key graphical features—zeros, asymptotes, and multiplicities. Mastering the following expressions allows you to translate any set of constraints into a function.

GENERAL POLYNOMIAL IN FACTORED FORM

p(x) = a(x − r₁)^m₁ (x − r₂)^m₂ ⋯ (x − rₖ)^mₖ

Here a is the leading coefficient, r₁, r₂, …, rₖ are the distinct real zeros, and m₁, m₂, …, mₖ are their respective multiplicities. The degree of the polynomial is n = m₁ + m₂ + ⋯ + mₖ. If mᵢ is odd, the graph crosses the axis at x = rᵢ; if mᵢ is even, it touches and rebounds.

GENERAL RATIONAL FUNCTION

r(x) = a · N(x) / D(x) = a · [(x − z₁)(x − z₂) ⋯] / [(x − p₁)(x − p₂) ⋯]

N(x) and D(x) are polynomials in factored form. Zeros of N(x) give x-intercepts; zeros of D(x) produce vertical asymptotes (if the factor does not cancel) or holes (if it does cancel with a factor in N(x)).

HORIZONTAL / SLANT ASYMPTOTE RULES

deg(N) < deg(D) → y = 0 ; deg(N) = deg(D) → y = aₙ / bₙ ; deg(N) = deg(D) + 1 → slant asymptote

Where aₙ and bₙ are the leading coefficients of the numerator and denominator, respectively. These asymptotic behaviors describe the long-run (end) behavior of rational models and are critical when interpreting limiting values in applied contexts.

SOLVING FOR THE LEADING COEFFICIENT

a = y₀ / [(x₀ − r₁)^m₁ (x₀ − r₂)^m₂ ⋯ (x₀ − rₖ)^mₖ]

Given a known point (x₀, y₀) that is not a zero, substitute into the factored form and solve for a. The y-intercept (0, y₀) is often the most convenient choice when it is provided.

💡 AP Exam Tip

On the AP Precalculus exam, you will frequently be asked to determine the form of a function from a description of its zeros, asymptotes, and one additional condition. Always write the factored form first, then use the extra condition to solve for a. Never try to expand and match coefficients unless the problem specifically requires standard form.

SECTION 5

Classifying Model Behavior — Polynomial vs. Rational

One of the most important skills in model construction is deciding whether a polynomial or a rational function is the appropriate choice. The decision hinges on the behavior described in the problem: polynomial models are suitable when the domain is all real numbers and the output grows without bound, while rational models are needed when the situation involves limiting values, undefined inputs, or quantities that approach but never reach a boundary. The diagram below contrasts these two families side by side.

The left panel shows a polynomial with continuous, unbounded behavior and x-intercepts at its zeros. The right panel shows a rational function with a vertical asymptote (VA) and a horizontal asymptote (HA), illustrating the restricted domain and limiting output behavior that distinguish rational models.

Comparison of key features guiding the choice between polynomial and rational models.

Feature	Polynomial Model	Rational Model
Domain	All real numbers	All reals except where D(x) = 0
Continuity	Always continuous	Discontinuous at vertical asymptotes and holes
End Behavior	Output → ±∞ as x → ±∞	Output → finite limit (HA) or grows like a polynomial (slant/no HA)
When to Use	Modeling quantities with no natural ceiling or floor; projectile motion, revenue	Modeling quantities approaching a limit; concentration, efficiency, average cost

SECTION 6

Worked Example — Constructing a Rational Model

A chemical mixing problem states: a large tank initially contains 50 liters of pure water. A brine solution containing 3 grams of salt per liter is pumped in at a rate of 2 liters per minute. The concentration C(t) of salt in the tank at time t minutes is modeled by a rational function. The well-mixed solution is not drained, so the total volume increases. Construct C(t) and determine the limiting concentration as t → ∞.

Constructing and Interpreting C(t)

Step 1 — Identify the quantitiesThe salt entering the tank accumulates at 3 g/L × 2 L/min = 6 grams per minute. After t minutes, the total salt in the tank is 6t grams (since it started with 0 g). The total volume of liquid at time t is 50 + 2t liters.

Salt = 6t grams; Volume = (50 + 2t) liters

Step 2 — Write the rational modelConcentration equals total salt divided by total volume. Therefore, the rational function model is C(t) = 6t / (50 + 2t), where the numerator is a polynomial of degree 1 and the denominator is also degree 1.

C(t) = 6t / (50 + 2t)

Step 3 — Simplify and identify featuresFactor a 2 from the denominator: C(t) = 6t / [2(25 + t)] = 3t / (25 + t). The zero of the numerator is t = 0 (the concentration starts at 0 g/L). The zero of the denominator is t = −25, which is outside the physical domain t ≥ 0, so there is no vertical asymptote in the relevant domain.

C(t) = 3t / (25 + t), domain: t ≥ 0

Step 4 — Determine the horizontal asymptoteSince deg(numerator) = deg(denominator) = 1, the horizontal asymptote is the ratio of leading coefficients: y = 3 / 1 = 3. This means that as time grows without bound, the concentration approaches 3 grams per liter—the concentration of the incoming brine. This makes physical sense: eventually, the tank's content approaches the concentration of the input.

Horizontal asymptote: y = 3 g/L

Step 5 — Validate with a specific valueAt t = 25 minutes, C(25) = 3(25) / (25 + 25) = 75 / 50 = 1.5 g/L. At t = 100, C(100) = 300 / 125 = 2.4 g/L. The values are increasing toward 3 but never reaching it, confirming the asymptotic behavior and the model's correctness.

C(25) = 1.5 g/L, C(100) = 2.4 g/L — consistent with HA at y = 3

SECTION 7

Strengths and Limitations of Polynomial and Rational Models

No model is perfect, and understanding the tradeoffs between polynomial and rational models is essential for selecting the right tool. Polynomial models offer simplicity and differentiability everywhere, but they can wildly overshoot or undershoot outside the range of the data. Rational models capture asymptotic behavior and discontinuities elegantly, but they introduce domain restrictions and can be algebraically cumbersome. The table below summarizes these tradeoffs.

Strengths (✓) and limitations (✗) of polynomial versus rational function models.

Criterion	Polynomial Strength / Limitation	Rational Strength / Limitation
Simplicity	✓ Straightforward algebra; easy to differentiate and integrate	✗ Partial fractions or other techniques may be needed for algebraic manipulation
Extrapolation	✗ High-degree polynomials oscillate wildly outside data range (Runge's phenomenon)	✓ Asymptotes provide a natural long-term prediction
Modeling Limits	✗ Cannot model a quantity approaching a finite ceiling or floor	✓ Horizontal asymptotes naturally represent carrying capacity, limiting concentration, etc.
Domain	✓ Defined for all real inputs—no need to check for excluded values	✗ Must identify and exclude values where the denominator is zero
Data Fit (Interpolation)	✓ n + 1 data points uniquely determine a degree-n polynomial	✓ Can match both intercepts and asymptotic constraints simultaneously

✦ KEY TAKEAWAY

Choosing between a polynomial and a rational model is like choosing between a telescope and a microscope: both reveal detail, but each is optimized for a different kind of observation. Use a polynomial when the domain is unrestricted and the output grows freely; use a rational function when the system naturally plateaus, has forbidden inputs, or exhibits asymptotic behavior. On the AP exam, context clues like "approaches a limit" or "undefined at x = k" are strong signals for a rational model.

SECTION 8

Connections to Calculus and Advanced Modeling

The model-construction techniques you develop in this unit are not isolated AP Precalculus skills—they form the algebraic backbone for the analysis of derivatives, integrals, and limits in calculus. Understanding how a polynomial's degree governs its turning points directly foreshadows the relationship between a function and its first derivative. Similarly, the asymptotic analysis of rational functions is a gateway to formal limit computations and L'Hôpital's Rule. The table below maps Precalculus model-building concepts to their calculus counterparts.

Mapping precalculus model-building skills to their calculus applications.

Precalculus Concept	Calculus Extension
Zeros and factored form of p(x)	Finding critical points by factoring f′(x) = 0
Multiplicity and local behavior at a zero	First and second derivative tests for extrema and inflection points
Horizontal asymptotes of rational functions	Limits at infinity: lim(x→∞) r(x)
Vertical asymptotes and holes	One-sided limits, removable vs. non-removable discontinuities
End behavior determined by degree and leading coefficient	Dominance analysis and growth rate comparisons (polynomial vs. exponential)

Beyond calculus, polynomial and rational models appear throughout applied mathematics and the sciences. In engineering, transfer functions in control theory are rational functions whose poles (denominator zeros) determine system stability. In pharmacokinetics, rational models describe how drug concentration in the bloodstream rises and then decays toward an asymptote. Mastering model construction now equips you with a transferable analytical framework that will serve you across STEM disciplines.

SECTION 9

Practice Problems

A function f has zeros at x = −3, x = 1 (with multiplicity 2), and x = 5. The graph of f passes through (0, 30). Which of the following is a correct expression for f(x)?

A rational function g has a vertical asymptote at x = 4, a horizontal asymptote at y = 2, and an x-intercept at x = −1. Which of the following could be g(x)?

A polynomial p(x) has degree 4, a positive leading coefficient, zeros at x = −2 and x = 3 (each with multiplicity 1), and a repeated zero at x = 0 with multiplicity 2. The graph of p passes through (1, −18). Write p(x) in factored form and determine p(−1).

PROBLEM 4 — APPLIED

A company's average cost per unit (in dollars) for producing x units is modeled by the rational function A(x) = (5x + 8000) / x, where x > 0. (a) Identify the horizontal asymptote of A(x) and interpret it in the context of average cost. (b) Determine the number of units that must be produced so that the average cost per unit is $13. (c) Explain why A(x) is always greater than its horizontal asymptote for x > 0. (d) A second company has average cost B(x) = (5x² + 8000) / x². Compare the long-run behavior of A(x) and B(x) and explain which company achieves a lower per-unit cost at high production volumes.

PROBLEM 5 — CRITICAL THINKING

A rational function r(x) satisfies all of the following conditions: r has a zero at x = 2 with multiplicity 1; r has a vertical asymptote at x = −1; r has a hole at x = 3; the horizontal asymptote of r is y = −4. (a) Construct r(x) in fully factored form, including the leading coefficient. (b) State the domain of r. (c) Determine the y-intercept of r and the value of lim_{x→3} r(x).

SUMMARY

Lesson Summary

Constructing a function model begins with identifying the appropriate function family—polynomial for continuous, unbounded behavior, or rational for situations involving asymptotes, domain restrictions, and limiting values. The zeros and their multiplicities determine the factored structure, while the leading coefficient is found by substituting a known point. For rational models, the degrees of the numerator and denominator control the horizontal asymptote, zeros of the denominator produce vertical asymptotes or holes, and every feature must be validated against context.

The construction process follows a clear workflow: select the family, write the factored form encoding all known zeros and asymptotes, solve for the scaling constant a using an additional data point, and verify the model's end behavior, domain, and real-world reasonableness. These skills form the algebraic foundation for calculus-level analysis of limits, derivatives, and integrals, making them among the most transferable tools in your mathematical toolkit.