Elementary Functions, Equations & Graph Transformations

Elementary algebra is the interface contract for calculus. A derivative, integral, limit, or transform can be manipulated correctly only if the underlying expression has a known domain, its inverse operations are legitimate there, and the notation still denotes the same function after each rewrite.

The practical goal is not to memorize a separate trick for every equation. It is to maintain three objects together:

1. the function rule;
2. the allowed inputs;
3. the solution set or graph produced by the rule.

A symbolic line that forgets one of these can look plausible while changing the problem.

Why functions, equations, and graphs are three views of one model #

A real system often produces an output from an input. A function answers “what output follows from this input?”, an equation asks “which input produces the required output?”, and a graph shows all allowed input–output pairs at once. A rule is more reusable than a table of measurements, but only inside the domain where the model is intended to apply.

This is mathematical modeling in miniature: choose a family whose behavior matches the mechanism, fit its parameters, and keep the validity interval attached. The algebra can answer questions inside the model; it cannot prove that the chosen model describes the device outside the measured regime.

A function includes its domain #

A real-valued function is not merely a formula. It is a mapping $f:D\to\mathbb R$ with a specified domain $D$. The same expression can define different functions on different domains, and an inverse depends on both domain and range.

For a composite $f\circ g$, the usable inputs are

$$D_{f\circ g}=\{x\in D_g:g(x)\in D_f\}.$$

This rule is more dependable than collecting restrictions after a calculation. Apply it from the innermost operation outward:

The power expression deserves special care. Integer powers, rational powers, and the general real definition $u^v=e^{v\ln u}$ have different domains. Simplifying $(x^2)^{1/2}$ to $x$ is wrong over the reals; the principal square root gives $|x|$.

A small set of parent families #

Most elementary graphs are transformations or combinations of a few families. Their defining restrictions explain their characteristic geometry.

Polynomials and rational functions are algebraic. Exponential, logarithmic, and trigonometric functions are transcendental. That classification does not tell whether an equation has a closed-form solution: $x^5-x+1=0$ is polynomial but has no general expression by radicals, while $e^{2x}=7$ is solved immediately with a logarithm.

Read graph transformations from the input outward #

For

$$g(x)=a\,f\!\bigl(b(x-h)\bigr)+k,$$

the parameters have distinct geometric roles:

• $h$ moves the graph horizontally to the right by $h$;
• $b$ scales horizontal distances by $1/|b|$ and reflects across the vertical axis when $b<0$;
• $a$ scales vertical distances by $|a|$ and reflects across the horizontal axis when $a<0$;
• $k$ moves the graph vertically by $k$.

The reciprocal in the horizontal scale is the common trap. A point $(u,f(u))$ on the parent graph maps to

$$\left(h+\frac{u}{b},\;k+a f(u)\right).$$

The formula proves the transformation and also carries domain endpoints, asymptotes, intercept candidates, and special points.

For example,

$$g(x)=-2\sqrt{3(x-1)}+4$$

starts from $f(u)=\sqrt u$. Its domain condition $3(x-1)\ge0$ gives $x\ge1$. The parent endpoint $(0,0)$ maps to $(1,4)$; horizontal distances shrink by $1/3$, output values stretch by $2$ and reflect, and the graph is shifted upward by $4$.

Equations are statements about a solution set #

An equation $F(x)=G(x)$ asks for

$$S=\{x\in D:F(x)=G(x)\},$$

where $D$ is the common domain. Each algebraic operation must be classified:

• an equivalence step preserves exactly the same solution set;
• an implication-only step preserves every original solution but may add candidates;
• a restriction step may discard solutions and must be justified.

Adding the same expression to both sides is reversible. Multiplying by a nonzero expression is reversible only after proving it cannot vanish. Squaring, applying an even power, multiplying by a possibly zero denominator, or applying a non-injective function usually creates only an implication.

Consider

$$\sqrt{x+2}=x.$$

The left side is nonnegative, so any solution must satisfy $x\ge0$. Squaring gives

$$x+2=x^2, \qquad (x-2)(x+1)=0.$$

The candidates are $2$ and $-1$, but the sign condition already rejects $-1$. Substitution confirms $\sqrt4=2$, so $S=\{2\}$. The squared equation was a candidate generator, not an equivalent restatement.

Polynomial and rational problems are sign problems #

Factoring exposes the locations at which a product can change sign. For

$$R(x)=\frac{(x-2)^2(x+1)}{(x-3)(x+4)^3},$$

the critical points are zeros $-1,2$ and exclusions $-4,3$. A factor of odd multiplicity changes sign when crossed; an even multiplicity touches zero without changing sign. A sign chart therefore needs one test value in the rightmost interval and parity information at each critical point.

This method solves rational inequalities without expanding away their structure. Excluded denominator zeros remain open boundaries even when a common factor cancels. Numerator zeros can be included for $\le$ or $\ge$ only when the original expression is defined there.

Polynomial root formulas are less important than structural tools:

• the factor theorem connects $p(r)=0$ to a factor $(x-r)$;
• Vieta's relations connect roots to coefficients;
• division and multiplicity expose local sign behavior;
• end behavior follows from the leading term;
• exact algebra should be separated from numerical root approximation.

Exponential and logarithmic equations #

For $a>0$, $a\ne1$, the exponential $a^x$ is one-to-one and has inverse $\log_a x$. This makes

$$a^{u(x)}=a^{v(x)}\iff u(x)=v(x)$$

an equivalence, while

$$\log_a u(x)=\log_a v(x)\iff u(x)=v(x)>0$$

must retain both positivity conditions.

Logarithm laws are domain-sensitive identities. Over positive arguments,

$$\ln(uv)=\ln u+\ln v, \qquad \ln\!\left(\frac uv\right)=\ln u-\ln v.$$

Writing $\ln(x^2)=2\ln x$ for negative $x$ is invalid because the right side is undefined; the real identity is $\ln(x^2)=2\ln|x|$ for $x\ne0$.

If exponential terms cannot be reduced to a common base, logarithms may isolate the variable. The calculation still needs a sign check before taking a logarithm. An equation such as $3^x=-2$ has no real solution; applying $\ln$ to both sides is not a legal step.

Trigonometric equations live on a circle #

Trigonometric functions encode coordinates or ratios attached to an angle, so periodicity is part of the answer. Solving $\sin x=c$ means finding every angle with the required vertical coordinate, not only the principal inverse value.

For $|c|\le1$ and $\alpha=\arcsin c\in[-\pi/2,\pi/2]$,

$$\sin x=c \iff x=\alpha+2\pi k \quad\text{or}\quad x=\pi-\alpha+2\pi k, \qquad k\in\mathbb Z.$$

Equivalent compact parameterizations are acceptable only when they generate exactly the same set. Inverse trigonometric functions return principal values; they are not global inverses of periodic functions unless the original domain has been restricted to a one-to-one branch.

Substitutions such as $u=\sin x$ can reduce a trigonometric equation to an algebraic one, but each algebraic root must lie in the range $[-1,1]$, and every retained $u$ must then be lifted back to the full periodic family of angles.

A graph-first workflow #

Before detailed symbolic work, identify what the graph already forces:

1. determine the exact domain and boundary points;
2. identify symmetry and periodicity;
3. map zeros, exclusions, and easy reference points;
4. determine sign and end behavior;
5. carry parent-graph features through transformations;
6. use algebra to establish exact intersections or extrema;
7. verify candidates in the original rule.

This workflow also diagnoses impossible results. A claimed logarithmic solution outside the positive argument region, an odd number of intersections forced by a symmetry that creates pairs, or a rational graph crossing a vertical asymptote all signal that an algebraic step changed the problem.

Failure patterns to recognize #

• Cancelling through zero: $(x-1)/(x-1)=1$ only on the original domain $x\ne1$.
• Distributing a root: $\sqrt{u+v}$ is generally not $\sqrt u+\sqrt v$.
• Dropping absolute value: $\sqrt{x^2}=|x|$, not $x$ on all of $\mathbb R$.
• Reversing an inequality without a sign: multiplying by $q(x)$ requires splitting according to whether $q(x)$ is positive or negative.
• Using an inverse globally: $\arcsin(\sin x)=x$ only on the principal interval.
• Forgetting periodic copies: one principal trigonometric angle is rarely the complete solution.
• Trusting a plot as proof: finite resolution can hide holes, tangencies, close roots, or behavior outside the window.

The unifying habit is simple: transformations act on a function's rule and domain together, while equation manipulations act on a solution set. Keep those objects explicit and elementary algebra becomes a dependable base for every later chapter.