Precalculus

If functions have ever felt like a wall you could not get over, you are in the right place, and you are going to be fine. Here is something reassuring: you already use functions every day without calling them that. Drop money into a vending machine, press B4, and out comes one specific snack. The same button always gives the same snack. That is a function. We are just going to learn to write down something you already understand, one small step at a time, and no question is too small.

The big picture

A function is a machine with one job: take an input, give back exactly one output. That is the whole idea. It matters because almost everything you might want to predict, like a price from a quantity or a height from an age, is really one input going in and one answer coming out.

Key idea: one input, one and only one output. That word "only" is what makes it a function.

What a function is

A function is a rule that gives each input exactly one output. Picture the vending machine: the button is the input, the snack that drops is the output. The buttons you are allowed to press make up the domain (the legal inputs). The snacks that can actually come out make up the range (the outputs). The one rule you cannot break is that a single button never gives two different snacks.

Reading the notation out loud

We give a function a short name, usually the letter f, and we write f(x). Read that out loud as "f of x." It does not mean f times x. It means "the output of machine f when you feed in x." The little (x) is the input slot, not a multiplication.

Let us evaluate one together, slowly

Find f(3) when f(x) = 2x + 1.

1. Read the rule: 2x + 1 means "two times the input, then add one." (Just translating.)
2. The input is 3, so wherever we see x we gently put a 3: 2(3) + 1. (Filling the slot.)
3. Do the multiply first: 2 × 3 = 6. Now we have 6 + 1. (Multiplying comes before adding.)
4. Do the add: 6 + 1 = 7. (Last step.)

So f(3) = 7. That is the whole skill: take a rule, put in a number, read off one answer. Nice work.

Try it: with the same rule, what is f(5)?

Answer: Put 5 in the slot: 2(5) + 1 = 10 + 1 = 11. So f(5) = 11. See, you just did it again.

The vertical line test (a picture rule)

How can you look at a graph and tell if it is a function? Use the vertical line test: if any straight up-and-down line crosses the graph more than once, the graph is not a function. Here is why, in plain words: a vertical line stands over one input. If the graph is there twice, that one input has two outputs, which breaks the one-output rule. A U-shaped parabola passes the test. A full circle fails it, because a vertical line through the middle hits both the top and the bottom.

Finding the domain: which inputs are allowed

When a function is just a formula with no domain stated, the domain is every number that does not break the math. Only two things break it:

• You cannot divide by zero. Any input that makes the bottom of a fraction zero is banned.
• You cannot take the square root of a negative number and stay in real numbers. Whatever sits under a square root must be zero or bigger.

Worked example: domain of a fraction

Find the domain of f(x) = 1 / (x - 4).

1. Look only at the bottom: x - 4. (Fractions blow up when the bottom is zero.)
2. Ask when it equals zero: x - 4 = 0 at x = 4. (Solve the little equation.)
3. So 4 is the one banned input; everything else is fine.

Domain: all real numbers except 4. Read that as "any number you like, just not 4."

Worked example: domain of a square root

Find the domain of g(x) = √(x - 2). The symbol √ is read "the square root of."

1. Whatever is under the root must be zero or larger: x - 2 ≥ 0. (The sign ≥ is read "greater than or equal to.")
2. Add 2 to both sides: x ≥ 2. (We do the same thing to both sides to keep it balanced.)

Domain: all numbers 2 or bigger. And since a square root never returns a negative, the range here is all outputs 0 or bigger.

Try it: find the domain of f(x) = √(x + 5).

Answer: Need x + 5 ≥ 0, so x ≥ -5. Domain: all numbers −5 or bigger.

Writing a domain in interval notation

"All numbers 2 or bigger" is fine in a sentence, but textbooks and exams use interval notation, which is shorter once you can read it. A square bracket means "include this endpoint," a round bracket means "do not include it," and infinity always gets a round bracket because you never actually reach it.

• x ≥ 2 becomes [2, infinity). Read it as "from 2, including 2, on up forever."
• x > 2 becomes (2, infinity). Same thing, but 2 itself is out.
• All real numbers except 4 becomes (-infinity, 4) U (4, infinity). The U means "union," which is just "and also."
• -6 ≤ x with x ≠ 2 becomes [-6, 2) U (2, infinity).

Notice how the last one reads like a story: start at -6 and include it, run up to 2 but stop just short, skip 2, then continue past it forever. Every domain you write from now on can be said either way, and matching the two is a fast check that you understand your own answer.

Worked example: finding the range from a formula

Domains come from what breaks. Ranges are harder, because you have to ask what outputs are actually reachable. Three tactics cover most cases.

Case 1: a squared term. Find the range of f(x) = x² - 4x + 1. Complete the square to expose the lowest point.

1. Take half of -4, giving -2, and square it, giving 4. Add and subtract it: x² - 4x + 4 - 4 + 1.
2. Fold up the perfect square: (x - 2)² - 3.
3. Now read it. A square is never negative, so (x - 2)² is 0 at smallest, which happens at x = 2. Then f is at least 0 - 3 = -3.

Range: [-3, infinity). Check by plugging in: f(2) = 4 - 8 + 1 = -3. It matches.

Case 2: a square root. For g(x) = √(x - 2), the root symbol only ever returns values that are zero or positive, so the range is [0, infinity) no matter what the inside does.

Case 3: a fraction with a constant on top. For f(x) = 1 / (x - 4), ask which outputs are impossible. Could the output be 0? That would need 1 = 0, which never happens. Every other value is reachable. Range: (-infinity, 0) U (0, infinity).

The habit worth building: for the domain ask "what would break?", and for the range ask "what could never come out?"

Reading domain and range off a graph

On a graph, the domain is the shadow the curve casts on the x-axis (how far left and right it reaches), and the range is the shadow on the y-axis (how far down and up it reaches). Domain is the side-to-side story; range is the up-and-down story. For f(x) = x² - 1, the lowest point is at y = -1 and it climbs forever, so the range is all numbers −1 or bigger, while the domain is every real number (you can square anything).

When a formula has two restrictions

Sometimes a formula bans inputs for two reasons at once. The rule is simple: obey every ban, and keep only the numbers that survive all of them. Take h(x) = √(x + 6) / (x - 2). The root needs x + 6 ≥ 0, so x ≥ -6. The bottom cannot be zero, so x ≠ 2 (read "x is not equal to 2"). Keep both conditions: every number from −6 up, except 2.

Piecewise functions (different rules on different days)

A piecewise function uses one formula for some inputs and a different formula for others. This is not exotic: a garage that charges one rate for the first hour and another after that is piecewise. To use one, first check which piece your input belongs to, then use only that piece's formula.

Take f(x) = x + 1 when x < 0, and f(x) = x² when x ≥ 0. (Here < is read "less than.")

1. f(-2): is −2 less than 0? Yes, so use the first rule: -2 + 1 = -1.
2. f(3): is 3 at least 0? Yes, so use the second rule: 3² = 9.

Average rate of change (how fast, on average)

The average rate of change between two inputs is "how much the output changed, divided by how much the input changed": from x = a to x = b it is (f(b) - f(a)) / (b - a). It is the same idea as miles per hour, distance gained divided by time spent.

Worked example: average rate of change

Find it for f(x) = x² from x = 1 to x = 3.

1. Output at 3: f(3) = 3² = 9.
2. Output at 1: f(1) = 1² = 1.
3. Change in output: 9 - 1 = 8.
4. Change in input: 3 - 1 = 2.
5. Divide: 8 ÷ 2 = 4.

On average the output rose 4 for each step in x across that stretch. This small idea grows up into the derivative in calculus, so you are meeting something big early.

Where people get stuck

The most common stumble is reading f(x) as "f times x" and trying to multiply. It is not multiplication. f is the machine's name and (x) is the input slot. Whenever you see f(3), say to yourself "put 3 into machine f," and you will not mix it up. The second common stumble is declaring a domain "all real numbers" without checking. Before you answer, always scan for a fraction bottom that could be zero and a square root that could go negative.

Real-world reason this matters

Domains keep models honest. If C(n) is the cost of printing n posters, n must be a whole number that is zero or more, even if the formula would happily accept −3 or 2.5. If h(t) is the height of a ball t seconds after a throw, the model only makes sense from launch until it lands. The math and the real situation both get a vote in what the domain is.

Common misconceptions

• "f(x) means f times x." No. It is the output of f at input x.
• "Domain and range are the same." Domain is the inputs; range is the outputs. They usually differ.
• "Every formula allows all real numbers." Fractions ban a zero bottom; even roots ban a negative inside.
• "The horizontal line test decides if a graph is a function." That is the vertical line test's job; the horizontal test checks something else that comes later.
• "f(0) is always 0." f(0) is just the output when the input is 0. For f(x) = x² - 1 it is −1.

Recap

A function gives each input exactly one output. Read f(x) as "f of x," and to evaluate it, drop the number into the slot and simplify one step at a time. The domain is the legal inputs (never divide by zero, never take an even root of a negative), and the range is the outputs that actually come out. The vertical line test spots function graphs, piecewise functions switch rules at break points, and the average rate of change (f(b) - f(a)) / (b - a) measures how fast the output moved. You now hold the vocabulary the rest of the course is built on.

Sources

1. OpenStax. (2021). 1.1 Functions and function notation. In Precalculus 2e. openstax.org
2. OpenStax. (2021). 1.2 Domain and range. In Precalculus 2e. openstax.org
3. OpenStax. (2021). 1.3 Rates of change and behavior of graphs. In Precalculus 2e. openstax.org
4. OpenStax. (2021). 3.2 Domain and range. In Algebra and Trigonometry 2e. openstax.org
5. Dawkins, P. (n.d.). The definition of a function. Paul's Online Math Notes. Lamar University. tutorial.math.lamar.edu
6. Khan Academy. (n.d.). Precalculus. khanacademy.org
7. Math is Fun. (n.d.). What is a function? mathsisfun.com

If graphs that shift and flip and stretch have ever made your head spin, that is a normal reaction, and it is going to make sense by the end of this page. Here is the friendly secret: you are never learning a brand-new graph. You are taking one shape you already know and sliding it, flipping it, or stretching it, exactly like moving and resizing a photo on your phone. Same picture, new position. We will take it one tiny move at a time.

The big picture

A transformation is a predictable move applied to a graph you already know. Learn a handful of moves once, and you can draw dozens of new graphs without plotting point after point. That is a huge time saver, and it is why this topic is worth slowing down for.

Key idea: outside the function changes the output (up, down, taller). Inside the function changes the input (left, right, narrower), and inside moves feel backward.

Parent functions

A parent function is the plainest member of a family, such as y = x² for U-shaped parabolas or y = |x| for the V-shaped absolute value. Once you know the parent's shape, every relative is just that shape, moved or resized.

Shifts up and down (the friendly ones)

Starting from y = f(x):

• f(x) + k slides the graph up by k. (You are adding to every output, so every point rises.)
• f(x) - k slides it down by k.

These do exactly what you expect, because the change is happening to the output.

Shifts left and right (the surprising ones)

• f(x - h) slides the graph right by h. Yes, a minus moves it right.
• f(x + h) slides it left by h.

Read that twice, because it feels backward at first. The reason: the change is inside, happening to the input before the function acts. To land the same output you now need an x that is h larger, so the whole picture slides right. You do not have to love this yet, just remember: inside moves go opposite the sign.

Reflections (flips)

• -f(x) flips the graph across the x-axis (upside down). The minus is outside, so it flips the output.
• f(-x) flips the graph across the y-axis (left to right). The minus is inside, so it flips the input.

Stretches and squishes

Multiplying the whole function by a number a scales it up and down. If a > 1 the graph is a vertical stretch (taller, steeper). If a is between 0 and 1 it is a vertical compression (flatter). For example, y = 3x² is three times as tall as y = x² at every x.

Let us read one transformed function together, slowly

Describe g(x) = -2(x - 3)² + 4 compared to the parent y = x². Work from the inside out, one move per line.

1. See (x - 3) inside: shift right 3. (Inside move, opposite the sign.)
2. See the 2 multiplying: stretch vertically by 2. (Twice as tall.)
3. See the minus in front: flip across the x-axis, so the parabola now opens downward.
4. See + 4 outside: shift up 4.

The vertex started at (0, 0) and lands at (3, 4), opening down. Quick check: at x = 3, g(3) = -2(0)² + 4 = 4, matching the vertex. It checks out, so we did it right.

Track one point (the trick that makes it click)

Transformations feel abstract until you follow a single point. Suppose (4, 2) sits on y = f(x). Where does it go on y = 2 f(x - 3) + 1?

1. Inside x - 3 shifts right 3, so the input 4 becomes 4 + 3 = 7.
2. The outside 2 doubles the output, so 2 becomes 2 × 2 = 4.
3. The outside + 1 raises the output, so 4 becomes 4 + 1 = 5.

New point: (7, 5). Notice inside changes touched the x, outside changes touched the y. That one sentence organizes this entire topic.

Try it: how does y = |x + 2| - 5 move the parent y = |x|, and where is the new vertex?

Answer: + 2 is inside, so shift left 2; - 5 is outside, so shift down 5. The vertex moves from (0, 0) to (-2, -5). You just combined two moves.

Order matters: apply the moves in the right sequence

When several transformations stack up, doing them in the wrong order lands the graph in the wrong place. The safe order comes straight from the order of operations acting on the input and the output.

For y = a f(b(x - h)) + k, work like this:

1. Horizontal stretch or compression by a factor of 1/b (inside, so it is the reciprocal).
2. Horizontal shift right by h.
3. Vertical stretch by a, including a flip if a is negative.
4. Vertical shift up by k.

Worked example: why order matters. Compare y = 2x² + 3 with y = 2(x² + 3). They look almost identical on the page and behave completely differently.

• y = 2x² + 3: stretch first, then shift up 3. The vertex goes from (0, 0) to (0, 3).
• y = 2(x² + 3) = 2x² + 6: shift up 3 first, then stretch by 2 - which doubles the shift too. The vertex lands at (0, 6).

Three units of difference, purely from sequence. Check at x = 0: the first gives 2(0) + 3 = 3, the second gives 2(0 + 3) = 6. The arithmetic confirms it.

Worked example: a horizontal stretch, in detail

Horizontal scalings are the ones that trip people up, so here is one in full. Compare y = f(2x) with y = f(x).

Suppose f has a peak at x = 6. Where is the peak of f(2x)? Ask what input makes the inside equal 6: 2x = 6, so x = 3. The peak moved from 6 to 3 - the graph got squeezed toward the y-axis by a factor of 2.

Now y = f(x/2). Setting x/2 = 6 gives x = 12, so the peak stretched outward from 6 to 12.

The pattern: multiplying the input by a number bigger than 1 compresses horizontally; multiplying by a fraction stretches. It is the reverse of what the number suggests, for the same reason left and right shifts are reversed - the change happens to x before the function ever sees it. Whenever you doubt it, do what we just did: solve for the x that reproduces a known feature.

Even and odd functions (a kind of symmetry)

A function is even if f(-x) = f(x), which means its graph is a mirror image across the y-axis, like x². It is odd if f(-x) = -f(x), which means it has a spin-around symmetry through the origin, like x³.

Worked example: test even or odd, one line at a time

Classify f(x) = x³ - 4x. The test is to replace every x with −x and see what happens.

1. Substitute: f(-x) = (-x)³ - 4(-x). (Just swapping in −x.)
2. Simplify each piece: (-x)³ = -x³ and -4(-x) = +4x, so f(-x) = -x³ + 4x.
3. Factor out −1: -x³ + 4x = -(x³ - 4x) = -f(x).

Since f(-x) = -f(x), this function is odd. Most functions are neither even nor odd, and that is completely fine; these labels describe special symmetry, not every graph.

Build a formula from words

Write the equation for y = √x after it is shifted left 1, flipped across the x-axis, and shifted up 3.

1. Shift left 1: replace x with x + 1, giving y = √(x + 1). (Inside move, opposite the sign.)
2. Flip across the x-axis: multiply by −1, giving y = -√(x + 1). (Outside flip.)
3. Shift up 3: add 3, giving y = -√(x + 1) + 3. (Outside move.)

Check with a point: at x = 0, y = -√1 + 3 = -1 + 3 = 2. The match confirms the formula.

Where people get stuck

The number one trap is the direction of a horizontal shift. People see f(x - 3) and slide the graph left because of the minus. It actually slides right. Inside changes always run opposite their sign. The fix is a tiny ritual: whenever the change is inside the parentheses, do the opposite of what the sign suggests. The second trap is order: when a stretch and a shift are both present, do the stretch and any flip first, then the vertical shift last, or the shift gets scaled by mistake.

Real-world reason this matters

Transformations are the language of models. A shop whose daily profit follows P(t) but opens two hours later uses P(t - 2). Doubling every price stretches a revenue curve vertically by 2. A sound played at double speed is a horizontal squish of the original wave. Reading a formula as "the same shape, moved and scaled" is often the fastest way to understand a new situation.

Common misconceptions

• "f(x - 3) shifts left." It shifts right 3. Inside changes run opposite the sign.
• "f(x) + k and f(x + k) do the same thing." The first moves the graph up or down; the second moves it left or right.
• "-f(x) and f(-x) are the same flip." The first flips across the x-axis; the second flips across the y-axis.
• "Order never matters." It matters when a stretch or flip meets a shift; do stretches and flips before the vertical shift.
• "Every graph is even or odd." Most are neither. Symmetry is special.

Recap

Every transformed graph is a parent shape that has been moved or resized. Outside the function acts on outputs: f(x) + k shifts up or down, a f(x) stretches, -f(x) flips across the x-axis. Inside the function acts on inputs and runs opposite the sign: f(x - h) shifts right, f(bx) squishes horizontally, f(-x) flips across the y-axis. Track one known point through each move to check your work, and use f(-x) to test for even (matches f(x)) or odd (matches -f(x)) symmetry.

Sources

1. OpenStax. (2021). 1.5 Transformation of functions. In Precalculus 2e. openstax.org
2. OpenStax. (2021). 3.5 Transformation of functions. In Algebra and Trigonometry 2e. openstax.org
3. OpenStax. (2021). 1.6 Absolute value functions. In Precalculus 2e. openstax.org
4. OpenStax. (2021). 3.2 Quadratic functions. In Precalculus 2e. openstax.org
5. Dawkins, P. (n.d.). Transformations. Paul's Online Math Notes. Lamar University. tutorial.math.lamar.edu
6. Dawkins, P. (n.d.). Graphing functions. Paul's Online Math Notes. Lamar University. tutorial.math.lamar.edu
7. Khan Academy. (n.d.). Precalculus. khanacademy.org

Composition and inverses sound like heavy words, but the ideas behind them are ones you already live with. We will build the idea up from simple pieces. Composition is just doing two steps in a row, like a discount and then a tax. An inverse is just an undo button, like rewinding what you did. We will go slowly and check every step.

The big picture

Composition chains two functions so the answer from the first becomes the question for the second, like an assembly line. An inverse runs a function backward to recover the number you started with, like retracing your steps home. Both ideas show up constantly, so they are worth getting comfortable with now.

Key idea: composition means "do the inside first." An inverse means "undo, step by step, in reverse."

Composing functions

The composition of f and g is written f(g(x)) and read "f of g of x." It means: put x into g first, then take whatever comes out and put it into f. The inside function runs first. Order matters, so f(g(x)) is usually not the same as g(f(x)).

Let us evaluate one together, slowly

Let f(x) = x² + 1 and g(x) = 3x - 2. Find f(g(4)).

1. Do the inside, g, first: g(4) = 3(4) - 2. (g runs before f.)
2. Finish that arithmetic: 3(4) - 2 = 12 - 2 = 10.
3. Now feed 10 into the outside, f: f(10) = 10² + 1.
4. Finish: 100 + 1 = 101.

So f(g(4)) = 101. To get a general formula, put all of g into f: f(g(x)) = (3x - 2)² + 1.

Order really does matter

With the same two functions, compare both orders.

• f(g(x)) = (3x - 2)² + 1 = 9x² - 12x + 5.
• g(f(x)) = 3(x² + 1) - 2 = 3x² + 1.

These are different. Test at x = 1: the first gives 9 - 12 + 5 = 2, the second gives 3 + 1 = 4. So never assume the two orders agree.

The domain of a composition (the sneaky part)

A composition can be more restricted than either function alone, because the inside function's output has to be legal input for the outside one. Two rules apply at once:

1. x must be in the domain of the inside function g.
2. g(x) must be in the domain of the outside function f.

Worked example 1. Let f(x) = √x and g(x) = x - 3. Find the domain of f(g(x)).

• g accepts every real number, so rule 1 bans nothing.
• f needs its input to be zero or bigger, so we need g(x) ≥ 0, that is x - 3 ≥ 0, so x ≥ 3.
• Domain: [3, infinity). And indeed f(g(x)) = √(x - 3), which agrees.

Worked example 2, where the simplified formula lies. Let f(x) = x² and g(x) = 1/(x - 1). Then f(g(x)) = 1/(x - 1)², which looks like it should ban only x = 1. That happens to be right here, but watch what rule 1 does in general: even when the algebra tidies a restriction away, an x that g itself rejects is still banned.

Take f(x) = x² and g(x) = √x. Composing gives f(g(x)) = (√x)² = x, which as a formula accepts everything. But g rejects negatives, so the honest domain of the composition is [0, infinity), not all reals. Always find the domain before you simplify, because simplifying can quietly erase a restriction that is still real.

Inverse functions (the undo button)

An inverse function, written f⁻¹ and read "f inverse," undoes whatever f did. If f turns 3 into 7, then f⁻¹ turns 7 back into 3. A warning about the notation: that little −1 is not an exponent here and does not mean one over f. It is a name that says "the undo function."

An inverse exists only when f is one-to-one, meaning every output came from just one input. You can test this with the horizontal line test: if no flat line crosses the graph more than once, f is one-to-one.

Find an inverse, one step at a time

Find the inverse of f(x) = 2x + 6. The recipe is: write y, swap x and y, then solve for y.

1. Write it as y = 2x + 6. (Just renaming the output y.)
2. Swap x and y: x = 2y + 6. (Swapping is what "reverse the machine" means.)
3. Subtract 6 from both sides: x - 6 = 2y. (Undo the plus 6 first.)
4. Divide both sides by 2: (x - 6)/2 = y. (Undo the times 2.)

So f⁻¹(x) = (x - 6)/2. Notice we undid the operations in reverse order, like taking off your shoes before your socks would be wrong; you reverse the steps.

Check it with composition

A real inverse, composed with the original, must return x. Test it:

1. Start with f(f⁻¹(x)) = 2 · ((x - 6)/2) + 6. (Put the inverse into f.)
2. The 2 and the divide-by-2 cancel: (x - 6) + 6.
3. Simplify: x.

It came back to x, so the inverse is correct. That is a satisfying way to be sure you did not slip.

Strictly, a full verification checks both directions, because a function can undo another one way and not the other. So also do:

1. f⁻¹(f(x)) = ((2x + 6) - 6)/2. (Put f into the inverse this time.)
2. The +6 and -6 cancel: (2x)/2.
3. Simplify: x.

Both compositions returned x, so these two really are inverses. On an exam, showing both lines is what earns full credit.

Domains and ranges trade places

Because an inverse reverses the arrows, it also swaps the two sets: the domain of f is the range of f inverse, and the range of f is the domain of f inverse. This one fact answers questions that look hard.

Worked example. Take f(x) = √(x - 2), whose domain is [2, infinity) and range is [0, infinity). Find the inverse and state its domain and range.

1. Write y = √(x - 2) and swap: x = √(y - 2).
2. Square both sides: x² = y - 2.
3. Add 2: y = x² + 2. So f⁻¹(x) = x² + 2.

Now here is the catch. As a bare formula, x² + 2 accepts every real number. But the inverse's domain must equal f's range, which was [0, infinity). So the correct answer is f⁻¹(x) = x² + 2 for x ≥ 0, with range [2, infinity), matching f's domain.

Check it: f(6) = √4 = 2, and f⁻¹(2) = 4 + 2 = 6. It came home. Squaring is the step that introduces the extra branch, which is why the domain restriction has to be written by hand rather than trusted to the algebra.

Try it: for f(x) = 5x - 1, find f⁻¹(x) and then f⁻¹(9).

Answer: Swap and solve: x = 5y - 1 gives y = (x + 1)/5, so f⁻¹(x) = (x + 1)/5. Then f⁻¹(9) = 10/5 = 2. Well done.

The mirror picture

The graph of an inverse is the original graph reflected across the diagonal line y = x. That makes sense, because swapping x and y is exactly what flips a point across that diagonal. The point (3, 7) on f becomes (7, 3) on the inverse.

Restricting a domain to create an inverse

The parabola f(x) = x² is not one-to-one, because both 2 and −2 give 4, so a horizontal line hits it twice. But if we only keep the right half, x ≥ 0, it becomes one-to-one, and its inverse is √x. This trimming trick returns later when we define inverse trigonometric functions, so it is worth seeing once here.

Seeing a function as layers

Composition run backward is called decomposing, and it is a genuinely useful skill. The function h(x) = √(3x + 1) is really f(g(x)) with an inside g(x) = 3x + 1 and an outside f(x) = √x. Spotting the inside and the outside is exactly what the chain rule will ask of you in calculus, so practicing now pays off later.

Where people get stuck

The classic mix-up is reading f⁻¹(x) as one over f(x). It is not a reciprocal; it is the name of the undo function. If you ever need one over f(x), that is a fraction, a different thing entirely. The second common stumble is composing in the wrong order. Say it out loud: "f of g of x" means g happens first, the inside always goes first.

Real-world reason this matters

Composition chains real processes: a 20 percent discount followed by an 8 percent tax is t(d(p)) = 1.08(0.80p) = 0.864p. Inverses reverse processes: F(c) = (9/5)c + 32 turns Celsius into Fahrenheit, and solving backward gives c = (5/9)(F - 32), the inverse that turns Fahrenheit into Celsius.

Common misconceptions

• "f inverse of x means 1 over f(x)." No. The −1 names the inverse function, not a reciprocal.
• "f(g(x)) and g(f(x)) are the same." They usually differ; only special pairs, like a function and its inverse, agree.
• "Every function has an inverse." Only one-to-one functions do; others need a trimmed domain first.
• "The inverse graph is the original flipped upside down." It is the mirror image across the diagonal line y = x, not across the x-axis.
• "Composition is just multiplication." It is feeding one output into the next function, not multiplying.

Recap

Composition f(g(x)) feeds the output of g into f, so work inside out, and expect the two orders to differ. An inverse undoes f; it exists only when f is one-to-one (horizontal line test) and is found by swapping x and y and solving. Composing a function with its inverse returns x, and the inverse graph is the reflection of f across y = x. Seeing a function as an inside layer plus an outside layer sets you up for calculus.

Sources

1. OpenStax. (2021). 1.4 Composition of functions. In Precalculus 2e. openstax.org
2. OpenStax. (2021). 1.7 Inverse functions. In Precalculus 2e. openstax.org
3. OpenStax. (2021). 3.4 Composition of functions. In Algebra and Trigonometry 2e. openstax.org
4. OpenStax. (2021). 3.7 Inverse functions. In Algebra and Trigonometry 2e. openstax.org
5. OpenStax. (2021). 3.8 Inverses and radical functions. In Precalculus 2e. openstax.org
6. Dawkins, P. (n.d.). Inverse functions. Paul's Online Math Notes. Lamar University. tutorial.math.lamar.edu
7. Dawkins, P. (n.d.). Combining functions. Paul's Online Math Notes. Lamar University. tutorial.math.lamar.edu

If a graph with lots of bumps has ever looked like too much to read, you are going to be pleasantly surprised here. A polynomial graph is really just a smooth roller coaster, and a few simple clues tell you almost everything about its shape before you plot a single point. We are going to break it into small clues you can check one at a time.

The big picture

Two features control most of a polynomial's shape: how the arms point at the far left and right, and where the graph touches or crosses the x-axis. Learn to read those two things and you can sketch the graph quickly. That is the payoff of this lesson.

Key idea: the highest-power term decides the far-away behavior; the factors decide where the graph meets the x-axis.

The parts of a polynomial

A polynomial function is a sum of terms, each a number times x raised to a whole-number power, like P(x) = 2x³ - 5x² + x - 7. The biggest exponent is the degree, and the number in front of that biggest-power term is the leading coefficient. Here the degree is 3 and the leading coefficient is 2.

End behavior (which way do the arms point?)

End behavior is just where the graph heads as you look far to the right and far to the left. Only the leading term matters for this, and there are four cases:

Degree	Leading coefficient	Left arm	Right arm
even	positive	up	up
even	negative	down	down
odd	positive	down	up
odd	negative	up	down

A memory hook: even-degree graphs point both arms the same way, like a valley or a hill. Odd-degree graphs point their arms in opposite ways, like a line tilted across the page.

Zeros and x-intercepts

A zero (also called a root) is an input x that makes P(x) = 0. Every real zero is a place where the graph crosses or touches the x-axis. If a factor shows up more than once, the zero has a multiplicity equal to how many times it repeats, and that number tells you what the graph does there:

• Odd multiplicity: the graph crosses straight through the x-axis.
• Even multiplicity: the graph touches the axis and bounces back without crossing.

Let us read one graph together, slowly

Analyze P(x) = (x - 2)²(x + 1).

1. Find the degree: multiplying out, the top term is x · x · x = x³, so degree 3, leading coefficient positive. (Odd degree, positive: down on the left, up on the right.)
2. Find the zeros: x = 2 from (x - 2)², and x = -1 from (x + 1). (Set each factor to zero.)
3. Check each multiplicity: (x - 2)² is squared, so multiplicity 2, even, the graph touches and turns at x = 2. The factor (x + 1) appears once, multiplicity 1, odd, the graph crosses at x = -1.
4. Find the y-intercept: put in x = 0, giving P(0) = (-2)²(1) = 4, the point (0, 4).

Now the sketch draws itself: it rises from the lower left, crosses at x = -1, passes through (0, 4), dips down to touch at x = 2, and climbs to the upper right. See how much you knew before plotting anything.

Try it: for P(x) = -x³ + 4x, name the degree, the end behavior, and the zeros.

Answer: Degree 3 with a negative leading coefficient, so up on the left and down on the right. Factor: -x(x² - 4) = -x(x - 2)(x + 2), so the zeros are x = 0, 2, -2. Nicely done.

Smoothness and turning points

Two facts shape every polynomial. First, polynomials are smooth and continuous: no breaks, no holes, no sharp corners, ever. Second, a polynomial of degree n has at most n - 1 turning points, the spots where it changes from rising to falling or back. A degree-3 graph turns at most twice, which is exactly what our example did.

Build a polynomial from its zeros

Write a degree-3 polynomial, leading coefficient 1, with zeros −3 (once) and 1 (twice). Each zero c gives a factor (x - c), repeated for its multiplicity.

1. Set up the factors: P(x) = (x + 3)(x - 1)². (Zero −3 gives (x + 3); the double zero at 1 gives (x - 1) squared.)
2. Expand the square: (x - 1)² = x² - 2x + 1.
3. Multiply by (x + 3): the result is x³ + x² - 5x + 3.

Quick check: P(1) = 1 + 1 - 5 + 3 = 0 and P(-3) = -27 + 9 + 15 + 3 = 0. Both zeros check out.

A sign chart pins down where the graph lives

Zeros tell you where the graph meets the axis; a sign chart tells you which side of the axis it is on in between. Take the same P(x) = (x - 2)²(x + 1). Its zeros -1 and 2 split the line into three strips, so test one point in each.

Strip	Test x	Value	Position
x < -1	-2	(-4)²(-1) = -16	below the axis
-1 < x < 2	0	(-2)²(1) = 4	above
x > 2	3	(1)²(4) = 4	above

Read the pattern: the sign flips at x = -1 (odd multiplicity, so it crosses) and stays the same across x = 2 (even multiplicity, so it touches and turns back). The sign chart and the multiplicity rule confirm each other, which is exactly the kind of double check worth building into your habits.

Multiplicity three flattens as it crosses

Odd multiplicities all cross, but they do not cross the same way. Compare three graphs near x = 0:

• y = x (multiplicity 1) slices through the axis at a clean slant.
• y = x³ (multiplicity 3) flattens out right at the axis before continuing through, making a lazy S.
• y = x⁵ (multiplicity 5) flattens even more dramatically.

Check the flattening with numbers: at x = 0.1, the linear graph is at 0.1 while the cubic is at 0.001, a hundred times closer to the axis. Higher odd multiplicity means the graph lingers near the axis longer before committing to the crossing. The same is true for even multiplicities: y = x⁴ hugs the axis more tightly than y = x² before bouncing.

Worked example: build a polynomial through a given point

Zeros fix a polynomial's shape but not its vertical scale. One extra point pins down the leading coefficient.

Question: Find the degree-3 polynomial with a zero at x = -1, a double zero at x = 2, and passing through (0, 8).

1. Write the shape with an unknown scale factor: P(x) = a(x + 1)(x - 2)².
2. Use the given point. At x = 0: P(0) = a(1)(-2)² = 4a.
3. Set that equal to 8: 4a = 8, so a = 2.
4. Answer: P(x) = 2(x + 1)(x - 2)².

Verify the whole thing: P(-1) = 2(0)(9) = 0, P(2) = 2(3)(0) = 0, and P(0) = 2(1)(4) = 8. All three conditions hold, so the answer is right.

Finding a zero between two points

Because polynomials never jump, if P(a) is negative and P(b) is positive, the graph must cross the x-axis somewhere between a and b. For P(x) = x³ - x - 2: P(1) = -2 (below the axis) and P(2) = 4 (above it). A zero hides between 1 and 2. This is exactly how a calculator hunts down roots, by squeezing the interval tighter and tighter.

Where people get stuck

The most common error is expecting the graph to cross at every zero. It does not. At an even-multiplicity zero it only touches and turns back, like our x = 2 above. Before you decide cross or touch, look at whether the factor's power is odd or even. The other frequent slip is thinking the whole formula controls the far ends. It does not; only the single leading term does, so the middle terms cannot change which way the arms point.

Real-world reason this matters

Polynomials model quantities built from products. An open box cut from a 12 by 10 inch sheet by removing x-inch corner squares has volume V(x) = x(12 - 2x)(10 - 2x), a cubic whose sensible domain is 0 to 5 inches. Engineers use smooth polynomial curves for road and coaster transitions, and businesses fit low-degree polynomials to find the production level that maximizes profit near a turning point.

Common misconceptions

• "The degree equals the number of turning points." Degree n allows at most n - 1 turns; a graph may turn fewer times.
• "The graph crosses at every zero." At even-multiplicity zeros it touches and bounces back.
• "End behavior depends on the whole formula." Only the leading term controls the far ends.
• "A degree-n polynomial always has n real zeros." It has at most n; some may repeat or not be real.
• "A missing x-intercept is a mistake." A graph like x² + 1 never touches the x-axis, and that is perfectly legal.

Recap

A polynomial is a sum of whole-number-power terms. Its degree and leading coefficient set the end behavior (even degree, matching arms; odd degree, opposite arms; the sign points them up or down). Real zeros are x-intercepts, crossing at odd multiplicity and touching at even multiplicity. Graphs are smooth with at most n - 1 turning points, a sign change locates a zero between two inputs, and factored form lets you build or read a polynomial one zero at a time.

Sources

1. OpenStax. (2021). 3.3 Power functions and polynomial functions. In Precalculus 2e. openstax.org
2. OpenStax. (2021). 3.4 Graphs of polynomial functions. In Precalculus 2e. openstax.org
3. OpenStax. (2021). 5.3 Graphs of polynomial functions. In Algebra and Trigonometry 2e. openstax.org
4. OpenStax. (2021). 3.2 Quadratic functions. In Precalculus 2e. openstax.org
5. Dawkins, P. (n.d.). Graphing polynomials. Paul's Online Math Notes. Lamar University. tutorial.math.lamar.edu
6. Dawkins, P. (n.d.). Polynomial functions. Paul's Online Math Notes. Lamar University. tutorial.math.lamar.edu
7. Khan Academy. (n.d.). Precalculus. khanacademy.org

Finding the zeros of a cubic can feel like being handed a locked box with no key. Here is the reassuring news: there is a reliable key, and it is the same four steps every single time. If you have felt lost with this before, that was probably because no one showed you the steps in order. We will lay them out slowly, and you can lean on them.

The big picture

To solve a polynomial you find its zeros, and the whole job is a repeatable routine: list a short menu of candidates, test one, divide it out to shrink the problem, then factor what is left. Once you trust the routine, cubics become routine.

Key idea: a zero and a factor are two views of the same thing. If P(c) = 0, then (x - c) divides evenly.

Two theorems that save time

The remainder theorem says that when you divide P(x) by (x - c), the leftover remainder is exactly P(c). So you can find a remainder just by plugging in c, no long division needed. The factor theorem is the special case that matters most: (x - c) is a factor precisely when P(c) = 0.

Test a zero, slowly

Is x = 2 a zero of P(x) = x³ - 3x² + 4?

1. Put in 2: P(2) = 2³ - 3(2²) + 4. (Just substituting.)
2. Work it out: 8 - 12 + 4 = 0.
3. The result is 0, so yes, x = 2 is a zero and (x - 2) is a factor.

Divide it out with synthetic division

Once you know one factor, dividing by it shrinks the polynomial to something you can factor by hand. Synthetic division is a tidy shortcut for dividing by (x - c). Continuing the example, divide x³ - 3x² + 4 by (x - 2). First write the coefficients with a placeholder 0 for the missing x term: 1, -3, 0, 4. Use c = 2.

1. Bring the first number, 1, straight down.
2. Multiply 1 by 2 to get 2; add it to −3 to get −1.
3. Multiply −1 by 2 to get −2; add it to 0 to get −2.
4. Multiply −2 by 2 to get −4; add it to 4 to get 0. (That 0 is the remainder, confirming the factor.)

The numbers left, 1, -1, -2, are the quotient x² - x - 2. So P(x) = (x - 2)(x² - x - 2).

Finish the factoring

Factor the leftover quadratic: x² - x - 2 = (x - 2)(x + 1). So P(x) = (x - 2)²(x + 1), and the zeros are x = 2 (twice) and x = -1. See how dividing turned a cubic into an easy quadratic.

Long division, for when the divisor is not (x - c)

Synthetic division only works for divisors of the form (x - c). Polynomial long division works for any divisor, and it looks exactly like the long division you learned with numbers.

Worked example. Divide 2x³ + 3x² - 5 by x + 2. First insert a placeholder for the missing x term: 2x³ + 3x² + 0x - 5.

1. Divide the leading terms. 2x³ / x = 2x². That is the first term of the quotient.
2. Multiply and subtract. 2x²(x + 2) = 2x³ + 4x². Subtracting leaves -x² + 0x - 5.
3. Repeat. -x² / x = -x. Then -x(x + 2) = -x² - 2x, and subtracting leaves 2x - 5.
4. Repeat again. 2x / x = 2. Then 2(x + 2) = 2x + 4, and subtracting leaves -9.
5. Stop. The leftover -9 has lower degree than x + 2, so it is the remainder.

Result: 2x³ + 3x² - 5 = (x + 2)(2x² - x + 2) - 9.

Check it with the remainder theorem. Dividing by x + 2 is dividing by x - (-2), so the remainder should equal P(-2). Compute: 2(-8) + 3(4) + 0 - 5 = -16 + 12 - 5 = -9. It matches exactly, which confirms every subtraction above in one line of arithmetic. That check costs ten seconds and catches nearly every long-division slip.

Note also what the remainder tells you: it is not zero, so x + 2 is not a factor, and x = -2 is not a zero.

The rational zero test (your candidate menu)

When no zero jumps out, the rational zero test hands you a short list to try. Any rational zero is (a factor of the constant term) divided by (a factor of the leading coefficient). For 2x² - 3x - 2, the constant is −2 (factors 1, 2) and the leading coefficient is 2 (factors 1, 2), so the candidates are ±1, ±2, ±½. Testing x = 2 gives 0, a zero. This turns a hopeless search into a short checklist.

A full solve, start to finish

Find all zeros of P(x) = x³ - 6x² + 11x - 6.

1. List candidates: factors of −6 over factors of 1, so ±1, ±2, ±3, ±6.
2. Test one: P(1) = 1 - 6 + 11 - 6 = 0, so x = 1 is a zero and (x - 1) is a factor.
3. Divide it out: synthetic division with c = 1 on 1, -6, 11, -6 gives the quotient x² - 5x + 6.
4. Factor what is left: x² - 5x + 6 = (x - 2)(x - 3).

So P(x) = (x - 1)(x - 2)(x - 3), and the zeros are 1, 2, and 3. Candidates, test, divide, factor: that is the whole method, and it never changes.

Try it: show that x = -3 is a zero of P(x) = x³ + 2x² - 5x - 6, then name the other zeros.

Answer: P(-3) = -27 + 18 + 15 - 6 = 0, so −3 is a zero. Dividing out (x + 3) leaves x² - x - 2 = (x - 2)(x + 1), so the other zeros are 2 and -1.

How many zeros should you expect?

The fundamental theorem of algebra promises that a degree-n polynomial has exactly n zeros if you count multiplicity and allow complex numbers. Some zeros may not be real: x² + 9 = 0 has no real solutions, but it has two complex ones, 3i and -3i, where i² = -1. For real-coefficient polynomials, complex zeros always come in matching pairs, so if 3 + 2i is a zero, then 3 - 2i is too.

Worked example: building a polynomial from a complex zero

That conjugate-pair rule is more useful than it looks, because it lets you reconstruct a whole polynomial from partial information.

Question: Find a degree-3 polynomial with real coefficients, leading coefficient 1, whose zeros include 2 and 1 + i.

1. Fill in the missing zero. Real coefficients force the conjugate, so 1 - i is a zero too. That is all three.
2. Write the factors. P(x) = (x - 2)(x - (1 + i))(x - (1 - i)).
3. Multiply the complex pair first, because the imaginary parts cancel and leave something real. Group them as ((x - 1) - i)((x - 1) + i), which is a difference of squares: (x - 1)² - i². Since i² = -1, that becomes (x - 1)² + 1 = x² - 2x + 2.
4. Multiply by the real factor. (x - 2)(x² - 2x + 2) = x³ - 2x² + 2x - 2x² + 4x - 4 = x³ - 4x² + 6x - 4.

Check. P(2) = 8 - 16 + 12 - 4 = 0, as required. And every coefficient came out real, which is the sign that we handled the conjugate pair correctly. Had we used only 1 + i and skipped its partner, imaginary numbers would have survived into the final answer.

Always pair the conjugates and multiply them together first. Attacking the factors in the wrong order means dragging i through several lines of algebra for no reason.

Where people get stuck

The single most common error is the sign of c. To divide by (x + 3), you use c = -3, because x + 3 is the same as x - (-3). The plus sign fools people every time, so pause and rewrite it as a subtraction. The other frequent slip is forgetting the placeholder 0 for a missing power. If x³ - 3x² + 4 has no x term, you must still write a 0 in that spot, or every later column comes out wrong.

Real-world reason this matters

Zero-finding is how models get answered. Break-even analysis sets profit equal to 0 and factors. Setting the box volume from the last lesson equal to a target gives a cubic to solve. Engineers locate the natural vibration frequencies of a structure as the roots of a characteristic polynomial, using this same candidates-test-divide routine.

Common misconceptions

• "Dividing by (x + 3) means using 3." The form is (x - c), so (x + 3) means c = -3.
• "Synthetic division works for any divisor." It only handles linear divisors (x - c); use long division otherwise.
• "Skipping the placeholder 0 is harmless." Leaving out a missing power scrambles every later column.
• "No rational zero means no zeros." The rational test only lists rational candidates; irrational and complex zeros can still exist.
• "A remainder of 0 is a failure." The opposite: remainder 0 means the divisor is a factor.

Recap

The remainder theorem finds a remainder instantly, since dividing P(x) by (x - c) leaves P(c). The factor theorem turns zeros into factors and back. The standard solve is always the same four steps: list rational candidates (constant factors over leading factors), test until one gives 0, divide it out with synthetic division, and factor the smaller quotient. The fundamental theorem of algebra guarantees n zeros counting multiplicity over the complex numbers, with complex zeros of real polynomials arriving in matching pairs.

Sources

1. OpenStax. (2021). 3.5 Dividing polynomials. In Precalculus 2e. openstax.org
2. OpenStax. (2021). 3.6 Zeros of polynomial functions. In Precalculus 2e. openstax.org
3. OpenStax. (2021). 3.1 Complex numbers. In Precalculus 2e. openstax.org
4. OpenStax. (2021). 5.4 Dividing polynomials. In Algebra and Trigonometry 2e. openstax.org
5. Dawkins, P. (n.d.). Finding zeroes of polynomials. Paul's Online Math Notes. Lamar University. tutorial.math.lamar.edu
6. Dawkins, P. (n.d.). Dividing polynomials. Paul's Online Math Notes. Lamar University. tutorial.math.lamar.edu
7. Khan Academy. (n.d.). Precalculus. khanacademy.org

Rational functions come with words like asymptote, but the pictures behind them are simple. An asymptote is just a line the graph snuggles up to without ever quite touching, like a car easing toward a wall in a parking spot but stopping short. Let us meet the ideas one at a time and let the words become familiar.

The big picture

A rational function is one polynomial divided by another, and its personality comes from two questions: where does the bottom become zero, and how do the top and bottom degrees compare? Answer those, and you can sketch the graph. That is the entire plan.

Key idea: factor and cancel first. A leftover bottom factor makes a vertical asymptote; a canceled factor makes a hole.

What a rational function is

A rational function is a fraction of two polynomials, R(x) = P(x) / Q(x). Think of it as one machine's output divided by another's. The interesting behavior all comes from the bottom hitting zero and from comparing the sizes of the top and bottom.

Vertical asymptotes and holes

Set the bottom equal to zero to find the trouble spots. Then, after canceling any factor that appears on both the top and the bottom:

• A factor that survives only in the bottom gives a vertical asymptote, an invisible vertical wall the graph races up or down beside but never crosses.
• A factor that cancels from both top and bottom gives a hole, a single missing point in an otherwise smooth curve, like one erased pixel.

Let us find the vertical features together

Analyze R(x) = (x - 1) / ((x - 1)(x + 3)).

1. Spot the common factor: (x - 1) is on both top and bottom.
2. Cancel it, leaving 1 / (x + 3). (Cancelling is allowed everywhere except where it was zero.)
3. The canceled (x - 1) leaves a hole at x = 1.
4. The surviving bottom factor (x + 3) gives a vertical asymptote at x = -3.

Horizontal asymptotes by comparing degrees

The horizontal asymptote tells you the height the graph settles toward at the far left and far right. Just compare the top degree (call it n) with the bottom degree (call it d):

Comparison	Horizontal asymptote
top smaller (n < d)	y = 0 (the x-axis)
degrees equal (n = d)	y = the ratio of the leading coefficients
top bigger (n > d)	none (a slant or higher asymptote instead)

A full analysis, one step at a time

Analyze R(x) = (2x² + 1) / (x² - 4).

1. Factor the bottom: x² - 4 = (x - 2)(x + 2). Neither factor cancels, so there are vertical asymptotes at x = 2 and x = -2.
2. Compare degrees: both top and bottom are degree 2, so the horizontal asymptote is the ratio of leading coefficients, 2/1 = 2, that is y = 2.
3. Find the y-intercept: R(0) = 1 / (-4) = -¼.

So the graph hugs the line y = 2 far out, and breaks at the two vertical walls. A test point in each strip between the walls tells you whether that piece rides above or below y = 2.

A complete sketch, with a sign chart

Asymptotes give you the skeleton; a sign chart tells you which side of each asymptote the curve rides on. Here is the whole routine on R(x) = (x - 1) / ((x + 2)(x - 3)).

1. Vertical asymptotes. Nothing cancels, so the bottom's zeros give walls at x = -2 and x = 3.
2. x-intercept. A fraction is zero exactly when its top is zero (and the bottom is not), so x - 1 = 0 gives an x-intercept at (1, 0).
3. Horizontal asymptote. The top has degree 1, the bottom degree 2, so top is smaller and the asymptote is y = 0.
4. y-intercept. R(0) = (-1) / ((2)(-3)) = -1/-6 = 1/6.
5. Sign chart. The three critical x-values -2, 1, and 3 chop the number line into four strips. Test one point in each.

Strip	Test x	Value	Sign
x < -2	-3	(-4)/((-1)(-6)) = -0.67	below the axis
-2 < x < 1	0	1/6 = 0.17	above
1 < x < 3	2	(1)/((4)(-1)) = -0.25	below
x > 3	4	(3)/((6)(1)) = 0.50	above

Now describe the graph in words, precisely. Far to the left it creeps up toward y = 0 from below, then plunges to negative infinity as it approaches x = -2. It reappears from positive infinity just right of x = -2, falls through the y-intercept at 1/6, crosses the x-axis at (1, 0), and dives to negative infinity approaching x = 3. Past x = 3 it comes down from positive infinity and flattens toward y = 0 from above.

Notice the sign alternates at every critical value here, which happens because each factor appears exactly once. A repeated factor, such as (x + 2)², would let the sign stay the same across that wall instead of flipping.

A graph may cross a horizontal asymptote

Vertical asymptotes are genuine walls: the function is undefined there, so the graph can never touch one. Horizontal asymptotes are not walls at all - they only describe long-run behavior, and a graph is free to cross one in the middle.

Worked example. For R(x) = x / (x² + 1), the top has degree 1 and the bottom degree 2, so the horizontal asymptote is y = 0. But R(0) = 0/1 = 0, so the graph sits exactly on y = 0 at the origin. It crosses its own horizontal asymptote, then climbs a little, then settles back down toward it as x grows. Nothing is wrong; the asymptote was only ever a promise about the far left and far right.

Slant asymptotes

When the top degree is exactly one more than the bottom, the graph leans toward a tilted line, a slant asymptote. Find it by dividing; the quotient (ignore the remainder) is the line. For R(x) = (x² + 1)/(x - 1), dividing gives x + 1 with remainder 2, so the graph approaches y = x + 1, plus there is a vertical asymptote at x = 1.

Where exactly is a hole?

Find the hole of R(x) = (x² - 9)/(x - 3).

1. Factor the top: x² - 9 = (x - 3)(x + 3).
2. Cancel the shared (x - 3), leaving x + 3 for every x ≠ 3.
3. Find the height by putting x = 3 into the simplified form: 3 + 3 = 6.

The hole sits at (3, 6). A hole has both coordinates, and the y-coordinate always comes from the simplified expression, not the original.

Try it: find the vertical and horizontal asymptotes of R(x) = (3x) / (x - 5).

Answer: The bottom is zero at x = 5, so a vertical asymptote there. Both degrees are 1, so the horizontal asymptote is 3/1 = 3, that is y = 3. Great work.

Where people get stuck

The first common trap is calling every bottom-zero a vertical asymptote. If the factor cancels, it is a hole instead, so always factor before deciding. The second trap is treating a horizontal asymptote as an untouchable fence. A graph never crosses a vertical asymptote, but it absolutely can cross a horizontal or slant one in the middle region. The horizontal asymptote only promises what happens far to the left and right, not in the middle.

Real-world reason this matters

Rational functions describe averages and shared costs. If a batch of parts costs 500 dollars to set up plus 3 dollars each, the average cost per part is C(x) = (500 + 3x)/x = 500/x + 3. Its horizontal asymptote y = 3 is the long-run cost per part: the setup fee spreads thinner and thinner but never fully vanishes. Drug concentration after a dose and lens equations in optics read the same way.

Common misconceptions

• "Every bottom-zero is a vertical asymptote." If the factor cancels, it is a hole; factor first.
• "Graphs never touch any asymptote." Only vertical asymptotes are off-limits; horizontal and slant ones can be crossed in the middle.
• "The horizontal asymptote describes the whole graph." It describes only the far left and far right.
• "A hole shows up as a visible gap." It is a single missing point; software often cannot show it, so find it with algebra.
• "Equal degrees give y = 1." Equal degrees give the ratio of leading coefficients, which is 1 only when they happen to match.

Recap

A rational function is a ratio of polynomials. Factor and cancel first: canceled factors give holes (with heights from the simplified form), and surviving bottom zeros give vertical asymptotes. Compare degrees for the far ends: smaller top gives y = 0, equal degrees give the leading-coefficient ratio, and a top one degree bigger gives a slant asymptote found by dividing. Add the intercepts and one test point per strip for an accurate sketch.

Sources

1. OpenStax. (2021). 3.7 Rational functions. In Precalculus 2e. openstax.org
2. OpenStax. (2021). 5.6 Rational functions. In Algebra and Trigonometry 2e. openstax.org
3. OpenStax. (2021). 3.5 Dividing polynomials. In Precalculus 2e. openstax.org
4. Dawkins, P. (n.d.). Graphing rational functions. Paul's Online Math Notes. Lamar University. tutorial.math.lamar.edu
5. Dawkins, P. (n.d.). Rational expressions. Paul's Online Math Notes. Lamar University. tutorial.math.lamar.edu
6. Khan Academy. (n.d.). Precalculus. khanacademy.org
7. Math is Fun. (n.d.). Asymptote. mathsisfun.com

Exponential functions have a reputation for being wild and fast, but the idea underneath is simple: multiply by the same number again and again. If exponents have tripped you up before, that is a common experience, and this page starts from the ground up. You already understand doubling and half-off sales, and that is genuinely most of the concept.

The big picture

An exponential function grows or shrinks by a constant multiplier at each step, instead of adding a constant amount. That one difference, multiply versus add, is what makes it start slow and then take off (growth) or fade toward nothing (decay). It models money, populations, and medicine, so it is worth meeting gently.

Key idea: each step multiplies by the base. Bigger than 1 means growth; between 0 and 1 means decay.

The form of an exponential

An exponential function puts the variable up in the exponent: f(x) = a · bx, read "a times b to the power x." Here the base b is positive and not 1, and a is the starting value, since f(0) = a. This is a completely different creature from a power function like x², where the variable sits in the base instead.

Growth versus decay

• If b > 1, you have exponential growth: the output climbs, faster and faster. A friendly example base is 2 (doubling).
• If b is between 0 and 1, you have exponential decay: the output shrinks toward zero. A friendly example base is ½ (halving).

Every basic exponential graph passes through (0, a) and slides along the x-axis as a horizontal asymptote, getting closer and closer to y = 0 without ever landing on it.

Let us evaluate one together, slowly

Let f(x) = 3 · 2x.

1. f(0) = 3 · 20. Anything to the power 0 is 1, so this is 3 · 1 = 3.
2. f(2) = 3 · 22 = 3 · 4 = 12.
3. f(-1) = 3 · 2-1. A negative exponent means one over, so 2-1 = ½, giving 3 · ½ = 1.5.

Notice each step right multiplies the output by 2, and each step left divides by 2. That steady multiplier is the heartbeat of an exponential.

The natural base e

One base is so useful it earns its own letter: e, about 2.718. It shows up naturally in smooth, continuous growth like interest that compounds every instant. A function f(x) = ex behaves like any other growth exponential, just with this special base.

Turning a percent into a base

Real problems usually give a percent, and the base quietly packages it. Growing by rate r each step means multiplying by b = 1 + r; shrinking by rate r means b = 1 - r. A 6 percent raise gives base 1.06. Losing 15 percent of value each year gives base 0.85. Translating the percent into the base is almost always the first move.

A growth model, step by step

A colony of 200 bacteria doubles every hour, so P(t) = 200 · 2t. After 3 hours:

1. Put in t = 3: P(3) = 200 · 23.
2. Compute the power: 23 = 8.
3. Multiply: 200 · 8 = 1600 bacteria.

Doubling is growth because the multiplier (times 2) stays the same each hour, not the amount added.

A decay model, step by step

A medicine starts at 80 mg, and 25 percent leaves the body each hour, so 75 percent stays: A(t) = 80 · (0.75)t. After 2 hours: A(2) = 80 · 0.752 = 80 · 0.5625 = 45 mg. Because the base 0.75 is between 0 and 1, the amount fades toward zero without reaching it.

Compound interest and half-life

Money compounded n times a year follows A = P(1 + r/n)nt. Invest 1000 dollars at 8 percent compounded quarterly for one year: the quarterly rate is 0.08/4 = 0.02 over 4 periods, so A = 1000(1.02)4 ≈ 1082.43 dollars, a little more than plain 8 percent because interest earns interest. A half-life is the time for half of something to disappear. A 100 mg sample with a 6 hour half-life: after 18 hours that is 18/6 = 3 half-lives, so 100 → 50 → 25 → 12.5 mg.

Try it: 500 dollars grows by a factor of 1.1 each year, so V(t) = 500 · 1.1t. Find the value after 2 years.

Answer: V(2) = 500 · 1.12 = 500 · 1.21 = 605 dollars. That is exponential growth in action.

Worked example: building a model from two data points

Real problems rarely hand you a and b. They hand you two measurements, and you have to recover the model.

Question: A population is 18 thousand at year 2 and 486 thousand at year 5. Assuming exponential growth, find P(t) = a · bt.

1. Write both facts. a · b² = 18 and a · b⁵ = 486.
2. Divide the second by the first. The a cancels: b⁵ / b² = 486 / 18, so b³ = 27.
3. Solve for b. The cube root of 27 is 3, so b = 3. The population triples each year.
4. Back-substitute for a. a · 3² = 18, so 9a = 18 and a = 2.
5. Model: P(t) = 2 · 3t thousand.

Check both points. P(2) = 2 · 9 = 18. P(5) = 2 · 243 = 486. Both land exactly.

Step 2 is the move worth remembering: divide the equations. Dividing kills the unknown a and leaves a clean power of b. Subtracting would not have helped, because a is a multiplier, not an addend.

Continuous compounding and e

Compounding more often earns slightly more, but there is a ceiling. Take 1000 dollars at 8 percent for one year:

• Yearly: 1000(1.08)¹ = 1080.00 dollars.
• Quarterly: 1000(1.02)⁴ = 1082.43 dollars.
• Monthly: 1000(1 + 0.08/12)¹² = 1083.00 dollars.
• Continuously: 1000 · e0.08 = 1083.29 dollars.

Notice the gains shrink fast and then stop. Compounding every instant, the limit of the process, gives the formula A = P · ert, and this is precisely where the number e comes from. It is not an arbitrary constant someone chose; it is what (1 + 1/n)n settles down to as n grows without bound.

Exponential eventually beats linear

Linear growth adds the same amount each step; exponential multiplies. Compare y = 100x with y = 2x. At x = 5 the line leads (500 versus 32). By x = 10 they are close (1000 versus 1024). By x = 20 the exponential has run away (2000 versus over a million). Any growth exponential eventually passes any line, no matter how steep the line is. That eventual takeover is what "exponential" really means, not just "fast."

Reading the asymptote after a shift

Everything from the transformations lesson still works. The graph of y = 2x + 3 is the parent lifted up 3, so its horizontal asymptote rises to y = 3. And y = 2-x is a flip across the y-axis, turning growth into decay, which is why 2-x equals (½)x.

Where people get stuck

The most common misread is treating "exponential" as a synonym for "fast." Exponential decay actually shrinks, and growth can crawl at the start before it accelerates. The defining feature is the constant multiplier per step, not raw speed. The second stumble is confusing 2x with x². In 2x the variable is the exponent; in x² it is the base. Their long-run behavior is not even close, so keep track of where the x lives.

Real-world reason this matters

A 20,000 dollar car that loses 15 percent of its value a year follows V(t) = 20000 · (0.85)t, dropping to about 14,450 dollars after 2 years. It does not lose the same number of dollars each year; it loses 15 percent of an ever-smaller value, the signature of exponential decay. Population growth, radioactive dating, and savings accounts all run on the same idea.

Common misconceptions

• "Exponential just means fast." Decay shrinks, and growth can start slow; the real feature is a constant multiplier per step.
• "The base can be any number." The base must be positive and not 1; negative bases would swing wildly and are excluded.
• "x squared and 2 to the x are basically alike." One is a power function, the other exponential; their far behavior is utterly different.
• "Doubling adds the same amount each time." Doubling adds more each time; equal additions would be linear.
• "The graph eventually reaches the asymptote." It approaches the asymptote forever without touching it.

Recap

An exponential function f(x) = a · bx starts at a and multiplies by b each step: growth when b is bigger than 1, decay when b is between 0 and 1. Percents become bases through 1 + r or 1 - r, compound interest uses P(1 + r/n)nt, and half-life models use base ½ with the exponent counting half-lives. The graph passes through (0, a), hugs a horizontal asymptote, and eventually outruns any straight line.

Sources

1. OpenStax. (2021). 4.1 Exponential functions. In Precalculus 2e. openstax.org
2. OpenStax. (2021). 4.2 Graphs of exponential functions. In Precalculus 2e. openstax.org
3. OpenStax. (2021). 4.7 Exponential and logarithmic models. In Precalculus 2e. openstax.org
4. OpenStax. (2021). 6.1 Exponential functions. In Algebra and Trigonometry 2e. openstax.org
5. Dawkins, P. (n.d.). Exponential functions. Paul's Online Math Notes. Lamar University. tutorial.math.lamar.edu
6. Dawkins, P. (n.d.). Applications of exponentials and logarithms. Paul's Online Math Notes. Lamar University. tutorial.math.lamar.edu
7. Khan Academy. (n.d.). Precalculus. khanacademy.org

Logarithms confuse a lot of people, often because no one ever said plainly what they are. So here it is, plainly: a logarithm is just an exponent. It answers one question, "what power do I raise the base to?" If that word has ever made your stomach drop, you are in the right place, and by the end of this page it will feel much smaller.

The big picture

A logarithm undoes an exponential. Where an exponential asks "start at 1 and multiply by the base x times, what do I get?", a logarithm asks the reverse: "I ended up at this number, how many times did I multiply?" That reverse question is the whole idea.

Key idea: logb(y) = x and bx = y say the exact same thing. The log is the exponent.

The logarithm as a question

A logarithm is written logb(y) and read "log base b of y." The statement logb(y) = x means exactly bx = y. Because it undoes the exponential, the two functions cancel: logb(bx) = x and b raised to logb(y) gives back y.

Reading a couple out loud

log2(8) = 3, read "log base 2 of 8 is 3," because 23 = 8. And log10(1000) = 3 because 103 = 1000. Two bases are so common they get nicknames: the common log log(x) means base 10, and the natural log ln(x) means base e.

Convert between the two forms, slowly

Whenever a log confuses you, rewrite it as an exponential; the exponential form is usually easier to read.

1. log3(81) = 4 becomes 34 = 81. Check: 3 times 3 times 3 times 3 is 81.
2. Going the other way, 52 = 25 becomes log5(25) = 2.

The three laws of logarithms

Because logs turn multiplying into adding (their whole superpower), three friendly rules follow:

1. Product law: logb(MN) = logb(M) + logb(N). (A product inside becomes a sum outside.)
2. Quotient law: logb(M/N) = logb(M) - logb(N).
3. Power law: logb(Mp) = p · logb(M). (An exponent inside jumps out front.)

Expand a logarithm, one law at a time

Expand log2(8x³).

1. The inside is a product, so use the product law: log2(8) + log2(x³).
2. Evaluate the first piece: log2(8) = 3, since 23 = 8.
3. Bring the exponent down on the second piece with the power law: log2(x³) = 3 log2(x).

Result: 3 + 3 log2(x).

Condense a logarithm (the reverse)

Write 2 log(x) - log(y) as a single log.

1. Turn each coefficient into an exponent with the power law: log(x²) - log(y).
2. A difference of logs is a quotient, so use the quotient law: log(x² / y).

Try it: evaluate log2(32) and expand log(100x).

Answer: log2(32) = 5, since 25 = 32. And log(100x) = log(100) + log(x) = 2 + log(x). You just used the product law.

Why the laws are true (a one-line proof each)

These are not rules to memorize on faith. They are the exponent rules you already know, viewed from the other side.

Let M = bm and N = bn, so by definition logb(M) = m and logb(N) = n.

• Product law. MN = bm · bn = bm+n, so logb(MN) = m + n = logb(M) + logb(N). It is just "add the exponents."
• Quotient law. M/N = bm-n, so the log is m - n. It is "subtract the exponents."
• Power law. Mp = (bm)p = bmp, so the log is mp = p · logb(M). It is "multiply the exponents."

Read those three lines and you can see why there is no law for log(M + N): there is no exponent rule for bm + bn either. The gap in the log laws exactly mirrors a gap in the exponent laws.

The counterexample worth carrying around

If you are ever tempted to write log(a + b) = log(a) + log(b), test it with numbers you can do in your head, base 10:

• Left side: log(10 + 10) = log(20) ≈ 1.301.
• Right side: log(10) + log(10) = 1 + 1 = 2.
• 1.301 is not 2, so the rule is false.

Now test the law that is true: log(10 × 10) = log(100) = 2, and log(10) + log(10) = 2. They agree exactly. One thirty-second test settles which rule is real, and it works for every fake log identity you will ever be tempted by.

The change-of-base formula

Calculators usually only carry log (base 10) and ln (base e). To find a log in some other base, use the change-of-base formula: logb(x) = ln(x) / ln(b). For example, log5(20) = ln(20)/ln(5) ≈ 2.996/1.609 ≈ 1.86.

Here is where it comes from, in three lines. Let y = logb(x), which means by = x. Take the natural log of both sides: ln(by) = ln(x). The power law pulls y out front: y · ln(b) = ln(x). Divide: y = ln(x)/ln(b). Done.

Sanity check the answer. Is 1.86 reasonable for log5(20)? We know 5¹ = 5 and 5² = 25, so the answer must sit between 1 and 2, closer to 2 because 20 is closer to 25 than to 5. And 1.86 does exactly that. Bracketing between two easy powers is the fastest way to catch a keystroke error.

A few values worth knowing by heart

In every base b, logb(1) = 0 (because b0 = 1) and logb(b) = 1 (because b1 = b). And inverse pairs collapse at once: ln(e7) = 7, because the log just reads off the exponent.

Evaluate without a calculator

1. log4(1/16): ask "4 to what power is 1/16?" Since 42 = 16, we need 4-2 = 1/16, so the answer is −2.
2. log(0.001): base 10, and 0.001 = 10-3, so the answer is −3.

Fractions inside give negative logs. That is normal, not an error.

The graph and its domain

Since y = logb(x) undoes y = bx, its graph is the exponential curve reflected across the line y = x. It passes through (1, 0), climbs slowly forever, and has a vertical asymptote at x = 0, because you can only take the log of a positive number. A shift moves that wall: for y = log2(x - 3), the inside must be positive, so x - 3 > 0 gives domain x > 3.

Where people get stuck

The single biggest trap is inventing a law for the log of a sum. There is a product law, log(ab) = log(a) + log(b), but there is no rule that breaks up log(a + b). If you see a plus sign inside a single log, you are usually stuck for a good reason. The second common slip is mixing up ln and log: ln is base e (about 2.718), while a plain log means base 10. When in doubt, write out the base you mean.

Real-world reason this matters

When numbers span huge ranges, scientists measure the exponent instead. Earthquake magnitude is the log of an intensity ratio, so a magnitude 6 quake shakes 102 = 100 times harder than a magnitude 4. Decibels measure loudness with logs, and pH is the negative log of acidity, so lemon juice at pH 2 is 100,000 times more acidic than water at pH 7. Logs turn multiplication of intensities into simple addition of scale points.

Common misconceptions

• "log(a + b) = log(a) + log(b)." False. The product law is log(ab) = log(a) + log(b); there is no law for a sum.
• "log(a)/log(b) = log(a) - log(b)." False. The quotient law is about log(a/b); a quotient of two logs is the change-of-base form.
• "You can take the log of any number." The inside must be strictly positive; log(0) and log(-5) are undefined.
• "A negative log output means an error." Logs of numbers between 0 and 1 are negative by design.
• "ln means base 10." ln is base e; a plain log means base 10.

Recap

A logarithm is an exponent: logb(y) = x means bx = y, so the log undoes the exponential. Memorize logb(1) = 0 and logb(b) = 1, expand and condense with the product, quotient, and power laws, and reach any base with change of base ln(x)/ln(b). The graph passes through (1, 0) with a vertical asymptote where the inside hits zero, and logarithmic scales like magnitude, decibels, and pH compress giant ranges into small numbers.

Sources

1. OpenStax. (2021). 4.3 Logarithmic functions. In Precalculus 2e. openstax.org
2. OpenStax. (2021). 4.5 Logarithmic properties. In Precalculus 2e. openstax.org
3. OpenStax. (2021). 4.4 Graphs of logarithmic functions. In Precalculus 2e. openstax.org
4. OpenStax. (2021). 6.5 Logarithmic properties. In Algebra and Trigonometry 2e. openstax.org
5. Dawkins, P. (n.d.). Logarithm functions. Paul's Online Math Notes. Lamar University. tutorial.math.lamar.edu
6. Dawkins, P. (n.d.). Exponential and logarithm functions. Paul's Online Math Notes. Lamar University. tutorial.math.lamar.edu
7. Khan Academy. (n.d.). Precalculus. khanacademy.org

These equations look tricky because the unknown is stuck up in an exponent or trapped inside a log. There is a clean tool for each situation, and once you know which tool to reach for, the rest is ordinary algebra. If word problems about doubling money or decaying medicine have felt out of reach, this is exactly where they start to open up.

The big picture

There are only two moves to learn. If the unknown is in an exponent, take a logarithm of both sides to pull it down. If the unknown is inside a logarithm, rewrite the equation in exponential form to peel the log away. Everything else is the algebra you already know.

Key idea: a log pulls an exponent down to the ground; exponential form frees a variable trapped inside a log.

Two strategies, named plainly

To solve an exponential equation (variable in the exponent), take a log of both sides and use the power law. To solve a logarithmic equation (variable inside a log), rewrite it in exponential form. Pick the tool that matches where the unknown is hiding.

The easy case: matching bases

Solve 2x = 32.

1. Rewrite 32 as a power of 2: 32 = 25.
2. Now both sides are 2 to a power, so the exponents must match: x = 5.

This shortcut only works when both sides share a base, but when it does, it is the fastest route.

The general case: take a logarithm

Solve 3x = 20. There is no clean power of 3 equal to 20, so bring in a log.

1. Take the natural log of both sides: ln(3x) = ln(20). (Same operation to both sides.)
2. Use the power law to drop the exponent: x · ln(3) = ln(20). (This is why we logged.)
3. Divide both sides by ln(3): x = ln(20)/ln(3).
4. Estimate: ≈ 2.996/1.099 ≈ 2.73.

A logarithmic equation

Solve log2(x) = 5.

1. Rewrite in exponential form: log2(x) = 5 means x = 25.
2. Compute: x = 32. Since 32 is positive, it is a legal input to the log, so it works.

Try it: solve 5x = 125 and ln(x) = 0.

Answer: Since 125 = 53, the first gives x = 3. And ln(x) = 0 means x = e0 = 1. Two different tools, two clean answers.

Isolate the exponential first

When the exponential is wrapped in extra arithmetic, peel that off before you touch a log. Solve 5 · 2x = 40: divide both sides by 5 to get 2x = 8, and since 8 = 23, x = 3. Likewise for 7 + 3x = 34: subtract 7 first to get 3x = 27, so x = 3. Taking a log while the extra terms are still attached is the classic wrong first move, because there is no law for the log of a sum.

Watch for extraneous solutions

Logs only accept positive inputs, so after solving a log equation you must check that every answer keeps each log's inside positive. A value that fails is called an extraneous solution and gets thrown out.

Solve log(x) + log(x - 3) = 1.

1. Condense the left side with the product law: log(x(x - 3)) = 1.
2. Rewrite in exponential form, base 10: x(x - 3) = 101 = 10.
3. Expand and set to zero: x² - 3x - 10 = 0.
4. Factor: (x - 5)(x + 2) = 0, so x = 5 or x = -2.
5. Check each: x = 5 keeps both logs positive and works; x = -2 makes log(x) undefined, so it is extraneous.

The only real solution is x = 5. Skipping the check is how people lose points here, so always circle back and test.

When the variable is in exponents on both sides

Nothing changes: log both sides, then use the power law twice and gather the x terms.

Worked example. Solve 3x+1 = 5x.

1. Take the natural log of both sides: ln(3x+1) = ln(5x).
2. Power law on each side: (x + 1) ln 3 = x ln 5. Notice the whole exponent comes down, brackets and all.
3. Distribute: x ln 3 + ln 3 = x ln 5.
4. Collect the x terms on one side: ln 3 = x ln 5 - x ln 3 = x(ln 5 - ln 3).
5. Divide: x = ln 3 / (ln 5 - ln 3) ≈ 1.0986 / (1.6094 - 1.0986) = 1.0986 / 0.5108 ≈ 2.15.

Check. Left side: 33.15 ≈ 31.9. Right side: 52.15 ≈ 31.9. They agree, so the answer holds.

The move that matters is step 4. Once the logs are just numbers, this is a linear equation in x, and you solve it the way you have solved linear equations since Algebra I.

Equations with e, and with logs on both sides

Worked example: base e. Solve 50e0.03t = 200.

1. Isolate the exponential: divide by 50 to get e0.03t = 4.
2. Take ln of both sides. Since ln and e undo each other exactly, the left collapses: 0.03t = ln 4.
3. Compute: ln 4 ≈ 1.3863, so t = 1.3863 / 0.03 ≈ 46.2.

Use ln whenever the base is e, because ln(ek) = k in one step with no leftover coefficient.

Worked example: logs on both sides. Solve ln(x + 3) - ln(x - 1) = ln 5.

1. Condense the left with the quotient law: ln((x + 3)/(x - 1)) = ln 5.
2. Both sides are a single ln, so the insides must be equal: (x + 3)/(x - 1) = 5.
3. Cross-multiply: x + 3 = 5(x - 1) = 5x - 5.
4. Solve: 8 = 4x, so x = 2.
5. Check the domain. At x = 2: x + 3 = 5 > 0 and x - 1 = 1 > 0. Both logs are legal, so x = 2 survives.

Worked example: dating a sample from its half-life

Question: Carbon-14 has a half-life of 5730 years. A wooden artifact retains 25 percent of its original carbon-14. How old is it?

1. Write the decay model using half-lives: A/A₀ = (1/2)t/5730.
2. Set the ratio to 0.25: (1/2)t/5730 = 0.25.
3. Recognize that 0.25 = (1/2)², so the bases match and the exponents must too: t/5730 = 2.
4. Solve: t = 11 460 years.

Sanity check by counting halvings: 100 percent to 50 percent is one half-life, and 50 percent to 25 percent is a second, so two half-lives, which is 2 x 5730. The algebra and the counting agree. When the remaining fraction is not a tidy power of one half, you would take logs instead and divide, exactly as in the doubling-time example below.

Doubling time (a favorite application)

How long does money take to double at 6 percent compounded annually? We need (1.06)t = 2.

1. Take the natural log of both sides: ln((1.06)t) = ln(2).
2. Power law drops the exponent: t · ln(1.06) = ln(2).
3. Divide: t = ln(2)/ln(1.06) ≈ 0.6931/0.0583 ≈ 11.9 years.

The banker's rule of 72 estimates it as 72/6 = 12 years, pleasingly close. That rule works precisely because of this logarithm.

Where people get stuck

The most common error is grabbing a log too early, while extra terms are still attached to the exponential. Isolate the exponential term first, always. The second frequent trap is forgetting the domain check on log equations and keeping an answer that makes a log's inside negative or zero. Every log equation deserves a final look: does each solution keep every inside positive?

Real-world reason this matters

These equations answer "when" questions everywhere: when a population reaches a threshold, when a radioactive tracer decays below a safe level, when an investment hits a goal, and how old a fossil is (carbon dating solves a decay equation for time). Any time the unknown sits in an exponent, a logarithm is the tool that brings it down to earth.

Common misconceptions

• "Take the log of both sides immediately." Isolate the exponential first; log(a + b) cannot be split apart.
• "ln(20)/ln(3) equals ln(20/3)." No. A quotient of two logs is just a number, and it is the change-of-base form.
• "Every algebra solution is a real solution." Log equations need a domain check; discard any value making an inside nonpositive.
• "Exponential equations always give whole numbers." Most need logs and give decimals like ln(20)/ln(3); the matching-base case is the special one.
• "You can divide to pull an exponent down." The exponent comes down only through logs (or matching bases), never by dividing.

Recap

For exponential equations, isolate the exponential, then either match bases or take a log of both sides and use the power law. For logarithmic equations, condense to one log, rewrite in exponential form (or equate the insides when both sides are logs of the same base), solve, and always check for extraneous solutions. The very same recipes answer real "how long until" questions, from doubling money to drug decay.

Sources

1. OpenStax. (2021). 4.6 Exponential and logarithmic equations. In Precalculus 2e. openstax.org
2. OpenStax. (2021). 4.7 Exponential and logarithmic models. In Precalculus 2e. openstax.org
3. OpenStax. (2021). 6.6 Exponential and logarithmic equations. In Algebra and Trigonometry 2e. openstax.org
4. Dawkins, P. (n.d.). Solving exponential equations. Paul's Online Math Notes. Lamar University. tutorial.math.lamar.edu
5. Dawkins, P. (n.d.). Solving logarithm equations. Paul's Online Math Notes. Lamar University. tutorial.math.lamar.edu
6. Dawkins, P. (n.d.). Applications of exponentials and logarithms. Paul's Online Math Notes. Lamar University. tutorial.math.lamar.edu
7. Khan Academy. (n.d.). Precalculus. khanacademy.org

Trigonometry is a big topic, and radians can look like a secret code at first. Here is the friendly truth: an angle is just an amount of turn, and the unit circle is a single picture that organizes all of it. If trig has felt like memorizing random symbols before, we are going to build it up slowly from things you can actually see.

The big picture

Everything in this chapter lives on one circle of radius 1. Each angle points to one spot on that circle, and the coordinates of that spot are the cosine and the sine. Get comfortable with the picture and the rest of trig has a home to sit in.

Key idea: on the unit circle, the point an angle lands on is (cosθ, sinθ), read "cosine theta, sine theta."

Two ways to measure a turn

You can measure an angle in degrees, where one full spin is 360, or in radians, where one full spin is 2π (about 6.28). A radian is the angle you get by laying one radius-length of arc along the rim, which is why radians connect so neatly to anything that spins. The bridge between the two systems is 180 degrees = π radians.

Converting between them

• Degrees to radians: multiply by π/180.
• Radians to degrees: multiply by 180/π.

Let us convert together, slowly

1. Convert 60 degrees to radians: 60 · π/180. The 60 and 180 reduce to 1/3, giving π/3.
2. Convert 3π/4 radians to degrees: (3π/4)(180/π). The π cancels, leaving 3 · 45 = 135 degrees.

Standard position and the unit circle

An angle is in standard position when its corner sits at the origin and its starting side lies along the positive x-axis. The unit circle is the circle of radius 1 centered at the origin. The spot where the angle's ending side meets that circle has coordinates (cosθ, sinθ). That one fact defines sine and cosine for every angle.

Reference angles (the shortcut to the whole circle)

A reference angle is the small acute angle between the ending side and the x-axis. It lets you turn any angle into a first-quadrant version you already know, then attach the right sign for the quadrant. For example, 150 degrees sits in Quadrant II with a reference angle of 30 degrees, so its values come from 30 degrees with signs adjusted for that quadrant.

The handful of special values to memorize

Knowing a few first-quadrant values unlocks the entire circle:

Angle	cos	sin
0	1	0
30 deg = pi/6	(root 3)/2	1/2
45 deg = pi/4	(root 2)/2	(root 2)/2
60 deg = pi/3	1/2	(root 3)/2
90 deg = pi/2	0	1

Where those exact values come from

You do not have to take that table on trust. Every entry falls out of two triangles you can draw from scratch.

The 45-45-90 triangle. Cut a square of side 1 along its diagonal. Both legs are 1, and by the Pythagorean theorem the diagonal is √(1² + 1²) = √2. Scale that triangle down so the hypotenuse is 1 (divide every side by √2) and the legs become 1/√2 = √2/2. Drop that triangle onto the unit circle at 45 degrees and its horizontal leg is the cosine and its vertical leg is the sine, so both equal √2/2 ≈ 0.707.

The 30-60-90 triangle. Take an equilateral triangle of side 1 and cut it down the middle. The half-triangle has hypotenuse 1, short leg 1/2 (half the base), and long leg √(1 - 1/4) = √(3/4) = √3/2. Place it on the unit circle at 30 degrees: the horizontal side is the long one, so cos 30° = √3/2 ≈ 0.866 and sin 30° = 1/2. Tip the same triangle onto its side for 60 degrees and the two swap: cos 60° = 1/2, sin 60° = √3/2.

Sanity check them against the picture. A small angle like 30 degrees lands near the right edge of the circle, so its x-coordinate should be large (0.866) and its y-coordinate small (0.5). A large angle like 60 degrees lands high up, so the opposite. If you ever misremember which is which, glance at the circle and ask "am I far right or far up?"

One more free check: because every point sits on a circle of radius 1, the Pythagorean theorem gives cos²θ + sin²θ = 1 for every angle. Test it on 30 degrees: (3/4) + (1/4) = 1. On 45 degrees: (1/2) + (1/2) = 1. Any pair of values that fails this test is wrong, no exceptions.

Signs by quadrant

On the unit circle, cosine is the x-coordinate and sine is the y-coordinate, so the signs follow directly from which quadrant you are in.

Quadrant	cos (x)	sin (y)	tan (y/x)
I (0 to 90)	+	+	+
II (90 to 180)	-	+	-
III (180 to 270)	-	-	+
IV (270 to 360)	+	-	-

You never need to memorize this table separately. Just picture where the point sits and read off whether its x and y are left or right, up or down.

Worked practice. Evaluate three values using reference angle plus sign.

• sin(5π/4): that is 225 degrees, Quadrant III, reference 45 degrees. Sine is negative there, so the answer is -√2/2.
• cos(11π/6): that is 330 degrees, Quadrant IV, reference 30 degrees. Cosine is positive there, so the answer is √3/2.
• sin(-π/3): that is -60 degrees, which is coterminal with 300 degrees, Quadrant IV, reference 60 degrees. Sine is negative there, so the answer is -√3/2.

Coterminal angles

Angles that differ by a full spin land on the same ending side and are called coterminal. Add or subtract 360 degrees (or 2π) as many times as needed: 400 degrees is coterminal with 40 degrees, and −30 degrees is coterminal with 330 degrees. Coterminal angles have identical sine and cosine, which is exactly why trig values repeat every revolution.

Evaluate an angle outside the first quadrant

Find cos 210° and sin 210°.

1. Locate the quadrant: 210 degrees is 30 degrees past 180, so it is in Quadrant III.
2. Find the reference angle: 210 - 180 = 30 degrees.
3. Decide the signs: in Quadrant III both x and y are negative, so cosine and sine are both negative.
4. Read off the values: cos 210° = -√3/2 and sin 210° = -1/2.

Every angle on the circle reduces to a first-quadrant value plus a sign decision. That two-step habit, reference angle then sign, is the most reliable method there is.

Try it: convert 225 degrees to radians and give its reference angle.

Answer: 225 · π/180 = 5π/4 radians. It sits in Quadrant III, so the reference angle is 225 - 180 = 45 degrees. Nicely reasoned.

Where radians pay off: arc length and area

For a circle of radius r and a central angle θ measured in radians:

• Arc length: s = rθ.
• Sector area: A = ½ r²θ.

Worked example: radius 6 cm, central angle π/3. Arc length is 6 · π/3 = 2π ≈ 6.28 cm. Sector area is ½(36)(π/3) = 6π ≈ 18.85 square cm. Both formulas need radians; plugging in degrees gives nonsense, which is the practical reason radians are the default in calculus and physics.

Where people get stuck

The number one trap is calculator mode. Degree mode and radian mode give completely different answers for the same keystrokes, so glance at the mode before you evaluate anything. The second trap is measuring the reference angle to the wrong axis. It is always measured to the x-axis, not the y-axis, and that is what keeps the memorized table values reusable.

Real-world reason this matters

Radians run anything that rotates: gear trains and bike sprockets trade angular speed between wheels of different sizes, and a point on a spinning rim of radius r moving at ω radians per second travels at v = rω. One degree of latitude is about 111 km because Earth's radius is about 6371 km and arc length is s = rθ. Whenever something spins, that little formula is nearby.

Common misconceptions

• "pi equals 180." pi radians equals 180 degrees; pi by itself is just the number 3.14159...
• "The calculator is always in the right mode." Degree versus radian mode silently changes every answer; check first.
• "Reference angles are measured to the y-axis." Always to the x-axis; that is what makes the table values reusable.
• "Coterminal angles are equal angles." They share an ending side but represent different amounts of turn, like 40 and 400 degrees.
• "Arc length works with degrees." s = r theta needs radians; using degrees inflates the answer by about 57 times.

Recap

Angles are measured in degrees (full circle 360) or radians (full circle 2 pi), linked by 180 degrees = pi radians. On the unit circle, an angle in standard position lands at (cos theta, sin theta). Reference angles plus quadrant signs evaluate any angle from a small memorized table, coterminal angles repeat values every revolution, and radian measure powers the formulas s = r theta, A = one-half r squared theta, and v = r omega.

Sources

1. OpenStax. (2021). 5.1 Angles. In Precalculus 2e. openstax.org
2. OpenStax. (2021). 5.2 Unit circle: sine and cosine functions. In Precalculus 2e. openstax.org
3. OpenStax. (2021). 7.1 Angles. In Algebra and Trigonometry 2e. openstax.org
4. OpenStax. (2021). 7.3 Unit circle. In Algebra and Trigonometry 2e. openstax.org
5. OpenStax. (2021). 5.4 Right triangle trigonometry. In Precalculus 2e. openstax.org
6. Khan Academy. (n.d.). Trigonometry. khanacademy.org
7. Math is Fun. (n.d.). Interactive unit circle. mathsisfun.com

Six trig functions can sound like six things to memorize, but really there are only two you need to know, and the other four are built from them. If names like secant and cotangent have felt like alphabet soup, that is completely understandable, and we are going to sort them into a tidy family you can keep straight.

The big picture

Sine and cosine are the two originals, straight from the unit circle. The other four are just quotients and reciprocals of those two. Learn how they are built once, and you never have to memorize them as separate mysteries.

Key idea: tangent is sine over cosine. The three with "co" or a reciprocal are one-over versions.

From two functions to six

Sine and cosine come from the unit circle. The rest are defined from them:

• Tangent: tanθ = sinθ / cosθ, read "sine over cosine."
• Cosecant: cscθ = 1 / sinθ.
• Secant: secθ = 1 / cosθ.
• Cotangent: cotθ = cosθ / sinθ.

Any of these is undefined wherever its bottom is zero. For instance, tangent is undefined at 90 degrees because cosine is zero there, and you cannot divide by zero.

Right triangles: SOH-CAH-TOA

For an acute angle in a right triangle, the old mnemonic SOH-CAH-TOA tells you which sides to use:

• Sine = Opposite / Hypotenuse.
• Cosine = Adjacent / Hypotenuse.
• Tangent = Opposite / Adjacent.

Let us read one triangle together, slowly

A right triangle has legs 3 and 4 and hypotenuse 5. For the angle whose opposite side is 3:

1. Sine is opposite over hypotenuse: 3/5 = 0.6.
2. Cosine is adjacent over hypotenuse: 4/5 = 0.8.
3. Tangent is opposite over adjacent: 3/4 = 0.75.

The size of the triangle does not matter, only the angle; a bigger triangle with the same angle gives the exact same ratios.

Signs by quadrant

Which functions are positive depends on the signs of x and y in that quadrant. A common memory aid is "All Students Take Calculus," naming what is positive in each quadrant:

Quadrant	Positive functions
I	All
II	Sine (and cosecant)
III	Tangent (and cotangent)
IV	Cosine (and secant)

All six from a point, step by step

The ending side of θ passes through (-3, 4). Find all six values.

1. Find the distance from the origin: r = √((-3)² + 4²) = √(9 + 16) = √25 = 5.
2. For a point, sine is y/r: sinθ = 4/5, so its reciprocal cscθ = 5/4.
3. Cosine is x/r: cosθ = -3/5, so secθ = -5/3.
4. Tangent is y/x: tanθ = 4/(-3) = -4/3, so cotθ = -3/4.

The point is in Quadrant II, and sure enough only sine and cosecant came out positive, matching "All Students Take Calculus."

Try it: in a right triangle with opposite 5, adjacent 12, hypotenuse 13, find sine, cosine, and tangent.

Answer: sinθ = 5/13, cosθ = 12/13, tanθ = 5/12. You matched the sides to the mnemonic perfectly.

Tangent at the special angles

Dividing the sine and cosine table gives tangent values worth knowing: tan 0 = 0, tan 45° = 1, tan 60° = √3, and tan 90° is undefined, because cos 90 = 0 puts a zero in the bottom. As the angle nears 90 degrees, tangent shoots up without bound, which matches the picture: a ramp at 89 degrees is nearly vertical, a huge rise for a tiny run.

A real measurement: angle of elevation

From a spot 50 meters from the base of a tower, the angle up to the top is 35 degrees. How tall is the tower?

1. The 50 m distance is the side next to the angle; the height h is opposite it.
2. Tangent links opposite and adjacent: tan 35° = h/50.
3. Solve for h: h = 50 · tan 35°.
4. Estimate: 50 · 0.700 = 35.0 meters.

The whole trick was choosing the right ratio: we knew adjacent and wanted opposite, and TOA points straight to tangent.

Worked example: solving a right triangle completely

"Solving a triangle" means finding every side and every angle. With one angle and one side you can get everything.

Question: In a right triangle, one acute angle is 35 degrees and the hypotenuse is 10. Find the other angle and both legs.

1. The other acute angle. The three angles total 180, and one is 90, so the acute pair totals 90. The other angle is 90 - 35 = 55 degrees.
2. The side opposite 35 degrees. Sine is opposite over hypotenuse, so sin 35° = opp / 10, giving opp = 10 sin 35° = 10(0.5736) ≈ 5.74.
3. The side adjacent to 35 degrees. Cosine is adjacent over hypotenuse, so adj = 10 cos 35° = 10(0.8192) ≈ 8.19.

Check with the Pythagorean theorem. 5.74² + 8.19² = 32.9 + 67.1 = 100.0, and the hypotenuse squared is 10² = 100. It closes, so the two legs are right.

A second sanity check costs nothing: 35 degrees is the smaller acute angle, so the side opposite it should be the shorter leg. It is (5.74 against 8.19). In any triangle, the bigger angle always faces the longer side.

Going backward: inverse trig finds the angle

When you know the sides and want the angle, use the inverse functions sin⁻¹, cos⁻¹, and tan⁻¹ (also written arcsin, arccos, arctan). They answer "what angle has this ratio?"

Worked example. A wheelchair ramp rises 2 metres over a horizontal run of 12 metres. What angle does it make with the ground?

1. Identify the ratio. Rise is opposite the angle, run is adjacent, so tangent is the one: tanθ = 2/12 = 0.1667.
2. Undo the tangent: θ = tan⁻¹(0.1667) ≈ 9.46 degrees.

Check by going forward again: tan(9.46°) ≈ 0.1666, which is the ratio we started from. Building the habit of reversing your own answer catches wrong-button errors instantly. (For context, accessibility codes typically cap ramps near 1 in 12, or about 4.8 degrees, so this ramp would be too steep.)

Worked example: all six functions from a point

Question: The terminal side of an angle in standard position passes through (-3, 4). Find all six trig functions.

1. Find r, the distance from the origin: r = √((-3)² + 4²) = √(9 + 16) = √25 = 5. Note r is always positive.
2. Use x = -3, y = 4, r = 5.

Function	Definition	Value
sin	y/r	4/5
cos	x/r	-3/5
tan	y/x	-4/3
csc	r/y	5/4
sec	r/x	-5/3
cot	x/y	-3/4

Check the signs against the quadrant. The point (-3, 4) is left and up, so Quadrant II, where sine is positive and cosine is negative. Our table agrees. And each reciprocal pair multiplies to 1: (4/5)(5/4) = 1, (-3/5)(-5/3) = 1, (-4/3)(-3/4) = 1. Three more free checks.

Cofunctions

In a right triangle the two acute angles add to 90 degrees, and the side opposite one is adjacent to the other. That symmetry gives the cofunction identities: sin(90° - θ) = cosθ and cos(90° - θ) = sinθ. This is literally why cosine is named "co-sine," the sine of the complementary angle. Check: sin 30 = 1/2 and cos 60 = 1/2.

Where people get stuck

The most common confusion is thinking sin⁻¹(x) means one over sine. On a calculator, sin⁻¹ is the inverse sine (arcsine), which gives back an angle; the reciprocal of sine is cosecant, a different thing. The second frequent trap is using SOH-CAH-TOA in a non-right triangle. Those ratios only hold in right triangles; for other triangles you will use the laws of sines and cosines later.

Real-world reason this matters

Right-triangle trigonometry is the surveyor's toolkit: tree and building heights from a single angle sighting, aircraft glide slopes, the grade of a road (a 6 percent grade is a tangent), and wheelchair ramp rules (rise over run is a tangent). GPS and cell networks pinpoint locations with these same ratios at scale.

Common misconceptions

• "sin inverse of x means 1/sin(x)." That is the inverse sine, which returns an angle; the reciprocal of sine is cosecant.
• "SOH-CAH-TOA works in any triangle." Only in right triangles; other triangles need the laws of sines and cosines.
• "The hypotenuse can be any side." It is always opposite the right angle and always the longest side.
• "Trig values depend on the triangle's size." They depend only on the angle; similar triangles share the same ratios.
• "Tangent is always defined." Tangent and secant blow up where cosine is 0; cotangent and cosecant fail where sine is 0.

Recap

Sine and cosine come from the unit circle; the other four are their quotients and reciprocals (tan = sin/cos, cot = cos/sin, sec = 1/cos, csc = 1/sin). In right triangles, SOH-CAH-TOA picks the ratio, and from a point (x, y) at distance r, sine is y/r, cosine is x/r, and tangent is y/x. Quadrant signs follow "All Students Take Calculus," cofunctions swap an angle with its complement, and tangent is undefined wherever cosine is zero.

Sources

1. OpenStax. (2021). 5.3 The other trigonometric functions. In Precalculus 2e. openstax.org
2. OpenStax. (2021). 5.4 Right triangle trigonometry. In Precalculus 2e. openstax.org
3. OpenStax. (2021). 7.2 Right triangle trigonometry. In Algebra and Trigonometry 2e. openstax.org
4. OpenStax. (2021). 7.4 The other trigonometric functions. In Algebra and Trigonometry 2e. openstax.org
5. OpenStax. (2021). 6.3 Inverse trigonometric functions. In Precalculus 2e. openstax.org
6. Khan Academy. (n.d.). Right triangles and trigonometry. In Trigonometry. khanacademy.org
7. Math is Fun. (n.d.). Solving triangles. mathsisfun.com

If graphs of waves have ever looked like a tangle of hills and valleys you could not pin down, you are in the right place, and this is going to make sense by the end. Here is the reassuring part: a sine or cosine graph is really just one wave shape that repeats forever, like ripples spreading across a pond at a steady beat. Once you learn four little dials that stretch and slide that single wave, you can draw any of these graphs. We will go slowly, one dial at a time, and no question is too small.

The big picture

Sine and cosine trace smooth repeating waves. Four numbers control the wave: how tall it is, how long one bump takes, where it starts sideways, and how high its center sits. Learn to read those four numbers and you can graph or model anything that cycles, like tides, daylight hours, or sound.

Key idea: one wave shape, four dials. Amplitude sets the height, the period sets the width of one cycle, the phase shift slides it sideways, and the midline sets the center.

The shape of one wave

A periodic function repeats the same values over and over. Sine starts in the middle, climbs to the top, comes back through the middle to the bottom, and returns, all within one period. Cosine is the very same wave, but it starts at the top. Their natural period is 2π, read "two pi" (about 6.28), the distance it takes to finish one full cycle before repeating.

The four dials

A general wave is written y = A sin(Bx − C) + D. Each letter is a dial you can turn:

• A is the amplitude: how far the wave rises above and dips below its center. We use the size of A, written |A| and read "the absolute value of A," because a height is never negative.
• B sets the period: the period is 2π ÷ |B|. A bigger B squeezes more waves into the same space, so the period gets shorter.
• C gives the phase shift: a sideways slide equal to C ÷ B.
• D is the midline: it lifts or lowers the center line to y = D.

Let us read one together, slowly

Find the amplitude, period, and midline of y = 3 sin(2x).

1. Look at A, the number in front: it is 3. So the amplitude is 3, and the wave reaches from −3 up to 3. (A sets the height.)
2. Look at B, the number multiplying x: it is 2. The period is 2π ÷ 2 = π. (One full wave now fits in a width of π instead of the usual 2π.)
3. Look for D, a number added at the very end: there is none, so D = 0. The midline is y = 0. (The center rests on the x-axis.)

So this is a wave 3 tall, centered on the x-axis, finishing a cycle every π. That is all three facts, read straight off the formula. Nice work.

A vertical shift, one step at a time

Read y = 2 cos(x) + 1.

1. Amplitude: the front number is 2, so the wave goes 2 above and 2 below its center. (Dial A.)
2. Midline: the + 1 at the end lifts the center up to y = 1. (Dial D.)
3. Top: center plus amplitude, 1 + 2 = 3. (The highest the wave gets.)
4. Bottom: center minus amplitude, 1 − 2 = −1. (The lowest the wave gets.)

Its period is 2π ÷ 1 = 2π, the usual length, and because it is cosine it starts at the top, at height 3, when x = 0.

Try it: state the amplitude, period, and midline of y = 4 sin(3x) − 2.

Answer: Amplitude 4 (the front number). Period 2π ÷ 3 (B is 3). Midline y = −2 (the −2 at the end). You just read all three dials at once.

Reading a graph backward

Often you are handed a wave and asked for its formula. That is the same reading in reverse. Suppose a wave has a highest point of 7, a lowest point of 1, and repeats every 4 units.

1. Midline is the average of top and bottom: (7 + 1) ÷ 2 = 4, so D = 4. (The center sits halfway between.)
2. Amplitude is half the gap from top to bottom: (7 − 1) ÷ 2 = 3, so A = 3. (Half of the full swing.)
3. The period is 4, and the period equals 2π ÷ B, so B = 2π ÷ 4 = π/2. (Solve for B.)

A cosine model is therefore y = 3 cos((π/2)x) + 4. Reading these three numbers off a picture is a common test question, and now it is just a three-step recipe.

The phase shift (sliding sideways)

The trickiest dial is C, the sideways slide. For y = 2 sin(2x − π/2), the shift is C ÷ B = (π/2) ÷ 2 = π/4 to the right. A reliable trick is to factor the inside so the slide shows itself: 2x − π/2 = 2(x − π/4), and the x − π/4 announces "slide right by π/4." The amplitude here is 2 and the period is 2π ÷ 2 = π.

Graphing with five key points

To draw one clean cycle, split the period into four equal steps and mark five landmark points. For plain sine: start on the midline going up, reach the top at one quarter, return to the midline at halfway, dip to the bottom at three quarters, and finish on the midline. For y = 3 sin(2x) the period is π, so each quarter step is π/4:

• (0, 0): on the midline, heading up.
• (π/4, 3): the top, because sin(π/2) = 1 and 3 × 1 = 3.
• (π/2, 0): back on the midline.
• (3π/4, −3): the bottom.
• (π, 0): one cycle complete.

Connect the five points with a smooth curve, then copy that cycle left and right as far as you need.

Reading the highest and lowest points

The top is always midline plus amplitude, D + |A|, and the bottom is midline minus amplitude, D − |A|. For y = 3 sin(x) + 2: top 2 + 3 = 5, bottom 2 − 3 = −1. For y = 4 cos(x) − 1: top 3, bottom −5. No graphing needed, just add and subtract.

Worked example: reading a real sinusoidal model

Question: Water depth in a harbour is modelled by h(t) = 3 sin((π/6)(t − 4)) + 7, where h is in metres and t is hours after midnight. Find the midline, amplitude, period, and the times and depths of the first high and low tides.

1. Midline. The + 7 outside lifts the whole wave, so the average depth is 7 m.
2. Amplitude. The multiplier 3 gives a swing of 3 m either way, so the depth ranges from 7 − 3 = 4 m to 7 + 3 = 10 m.
3. Period. The number multiplying t inside is π/6, and period is 2π divided by that: 2π / (π/6) = 12 hours. That is exactly right for a real tide, which cycles roughly twice a day.
4. Phase shift. The (t − 4) slides the wave 4 hours right, so the pattern that a plain sine would start at t = 0 instead starts at t = 4.
5. High tide. A sine peaks a quarter period after it crosses the midline going up. That crossing is at t = 4, and a quarter of 12 is 3, so high tide is at t = 7, that is 7:00 a.m. at 10 m.
6. Low tide. Half a period later: 7 + 6 = 13, so 1:00 p.m. at 4 m.

Verify the high tide algebraically. At t = 7, the inside is (π/6)(3) = π/2, and sin(π/2) = 1, so h = 3(1) + 7 = 10. Confirmed.

Every question a harbourmaster would actually ask - when is it deepest, when is it shallowest, how much does it vary, can my boat clear the bottom at 3 a.m. - is answered by those four numbers. That is why the four dials are worth naming precisely rather than treating the graph as a shape to memorize.

A quick word about tangent

Tangent behaves differently from its two cousins. Its natural period is π, not 2π; it has no amplitude because it climbs without limit; and it has vertical break lines, called asymptotes, wherever cosine is zero, at x = π/2 + πk for whole numbers k. Between two breaks it sweeps from far below zero, up through the middle, to far above.

Where people get stuck

The most common trip-up is the period. People see sin(4x) and think the period is 4. It is the opposite: the period is 2π ÷ 4 = π/2, because a bigger B packs the waves tighter. Whenever you see the number multiplying x, divide 2π by it rather than reading it off directly. The second common slip is forgetting that the midline moves the top and bottom with it: the maximum of A sin(x) + D is D + |A|, not just A.

Real-world reason this matters

Sinusoids describe almost everything that repeats: the hours of daylight across a year, the tide rising and falling in a harbor, the alternating current in a wall outlet (60 cycles a second in North America), and sound waves, where amplitude is loudness and the period sets the pitch. Fitting the midline, amplitude, and period to real measurements, exactly as we did when reading a graph backward, is how scientists turn a wiggly plot of data into a formula that predicts tomorrow.

Common misconceptions

• "The period of sin(Bx) is B." It is 2π ÷ |B|; a bigger B means a shorter period, not a longer one.
• "Amplitude is the whole distance from top to bottom." It is half that distance, measured from the midline.
• "The maximum of A sin(x) + D is A." It is D + |A|; the midline lifts the extremes with it.
• "In sin(Bx − C) the phase shift is C." It is C ÷ B; factor the inside to see the true slide.
• "A negative A makes the heights negative." A minus sign just flips the wave upside down; the amplitude is still |A|.

Recap

A sinusoid y = A sin(Bx − C) + D has four dials: amplitude |A| (height above the midline), period 2π ÷ |B| (width of one cycle), phase shift C ÷ B (sideways slide), and midline y = D (the center). The top sits at D + |A| and the bottom at D − |A|. Graph one cycle with five key points a quarter period apart, and build a model by reading the midline, amplitude, and period straight from the situation, choosing sine or cosine to match where the wave starts. You now hold every dial you need.

Sources

1. OpenStax. (2021). 6.1 Graphs of the sine and cosine functions. In Precalculus 2e. openstax.org
2. OpenStax. (2021). 6.2 Graphs of the other trigonometric functions. In Precalculus 2e. openstax.org
3. OpenStax. (2021). 7.6 Modeling with trigonometric functions. In Precalculus 2e. openstax.org
4. OpenStax. (2021). 8.1 Graphs of the sine and cosine functions. In Algebra and Trigonometry 2e. openstax.org
5. OpenStax. (2021). 1.5 Transformation of functions. In Precalculus 2e. openstax.org
6. Khan Academy. (n.d.). Trigonometry. khanacademy.org
7. Math is Fun. (n.d.). Amplitude, period, phase shift, and frequency. mathsisfun.com

If the phrase "trig identities" has ever made you want to close the book, you are in good company, and you are going to be fine here. An identity is nothing mysterious. It is just a fact about sine and cosine that stays true for every angle, a fact you can use to swap a messy expression for a tidy one, the way you might swap a long-winded phrase for a single word that means the same thing. We will meet them a few at a time and actually use each one before moving on.

The big picture

Identities are the trading rules of trigonometry: each one lets you replace an expression with an equal one. That is what makes a messy tangle of sines and cosines collapse into something short. It matters because every later step, solving equations, simplifying in calculus, adding waves in physics, leans on these trades.

Key idea: an identity is true for every angle, so you may swap one side for the other any time it helps.

What an identity is

An identity is an equation that holds for every angle where both sides make sense. Compare that with an ordinary equation like x + 2 = 5, which is true for only one value. The equation sin²θ + cos²θ = 1 is true for every θ, so it is a fact you can always lean on. One note on notation: sin²θ means (sinθ)², the sine value squared, read "sine squared theta."

The reciprocal and quotient trades

These just rename the three less-familiar functions in terms of sine and cosine.

• Reciprocal: csc = 1 ÷ sin, sec = 1 ÷ cos, cot = 1 ÷ tan. (Each is "one over" its partner.)
• Quotient: tan = sin ÷ cos and cot = cos ÷ sin.

A handy habit for almost any identity problem: when in doubt, rewrite everything in sines and cosines and watch what cancels.

The Pythagorean identity

The star of the show is sin²θ + cos²θ = 1. It comes straight from the unit circle, whose points satisfy x² + y² = 1 with x = cosθ and y = sinθ. Two more identities follow by dividing it through:

• 1 + tan²θ = sec²θ (divide the first by cos²θ).
• 1 + cot²θ = csc²θ (divide the first by sin²θ).

Memorize the first one cold. The other two you can always rebuild in a moment.

Let us use it together, slowly

If cosθ = 3/5 and θ sits in Quadrant I, find sinθ.

1. Start from the identity: sin²θ + cos²θ = 1. (Our always-true fact.)
2. Put in the known cosine: sin²θ + (3/5)² = 1. (Just substituting.)
3. Square the fraction: (3/5)² = 9/25, so sin²θ + 9/25 = 1.
4. Subtract from both sides: sin²θ = 1 − 9/25 = 16/25. (Undo the add.)
5. Take the square root: sinθ = 4/5. (Positive, because sine is positive in Quadrant I.)

One identity, five small steps, and the missing value appears. That is the whole rhythm of this topic.

Simplifying an expression

Simplify sinθ · cotθ.

1. Rewrite cot with the quotient trade: cotθ = cosθ ÷ sinθ. (Turn it into sine and cosine.)
2. So the product is sinθ × (cosθ ÷ sinθ).
3. The sine on top and the sine on the bottom cancel, leaving cosθ.

A cluttered product just became a single word, cosθ. See how the "rewrite in sines and cosines" habit paid off.

Proving an identity (transform one side)

Prove (1 − cos²θ) ÷ sinθ = sinθ. The rule for proofs: pick one side and transform it into the other. Never shuffle pieces across the equals sign as if you already believed it.

1. Work on the left. By the Pythagorean identity, 1 − cos²θ = sin²θ. (A rearranged form of the star identity.)
2. Substitute: sin²θ ÷ sinθ.
3. Cancel one sine: sinθ, which is exactly the right side.

Proved. The left side became the right side, one step at a time.

Try it: simplify cosθ · tanθ.

Answer: Write tan as sin ÷ cos: cosθ × (sinθ ÷ cosθ). The cosines cancel, leaving sinθ. Well done.

Two harder proofs, line by line

Real identity proofs need a strategy, not just substitution. Two strategies cover most of them: write everything in sines and cosines, and combine over a common denominator.

Proof 1: tanθ + cotθ = secθ cscθ. Start on the busier side, the left.

1. Rewrite both terms in sines and cosines: sinθ/cosθ + cosθ/sinθ.
2. Common denominator sinθcosθ: (sin²θ + cos²θ) / (sinθcosθ).
3. The top is the Pythagorean identity, which equals 1: 1 / (sinθcosθ).
4. Split the single fraction into a product: (1/cosθ)(1/sinθ).
5. Rename each reciprocal: secθ cscθ, which is the right side. Proved.

Check it numerically first, before you ever prove. At 30 degrees: tan 30 + cot 30 = 0.5774 + 1.7321 = 2.3094, and sec 30 csc 30 = 1.1547 x 2 = 2.3094. They match, so the identity is worth trying to prove. Had they disagreed, you would have saved yourself twenty minutes proving something false.

Proof 2: (1 + cosθ)/sinθ = sinθ/(1 - cosθ). Here neither side simplifies on its own, so use the conjugate trick.

1. Multiply the left side, top and bottom, by (1 - cosθ). That is multiplying by 1, so it is legal: [(1 + cosθ)(1 - cosθ)] / [sinθ(1 - cosθ)].
2. The top is a difference of squares: 1 - cos²θ.
3. Pythagorean identity turns that into sin²θ: sin²θ / [sinθ(1 - cosθ)].
4. Cancel one sine: sinθ / (1 - cosθ), which is the right side. Proved.

Where people get stuck on proofs: treating the identity like an equation and doing the same thing to both sides. You may not, because that assumes the very thing you are trying to establish. Work on one side alone until it turns into the other. If both sides are messy, it is acceptable to simplify each independently down to a common third expression, but never to move a term across the equals sign.

Even and odd angles

Flipping an angle to its negative reflects it across the x-axis, which flips the y-coordinate but leaves x alone. So sin(−θ) = −sinθ (sine is odd), cos(−θ) = cosθ (cosine is even), and tan(−θ) = −tanθ. Quick use: sin(−x) ÷ cos(−x) = −sin(x) ÷ cos(x) = −tan(x).

Sum and difference formulas

These unlock exact values for angles beyond the special few:

• sin(A ± B) = sinA cosB ± cosA sinB.
• cos(A ± B) = cosA cosB ∓ sinA sinB. (Notice the sign flips: a plus inside becomes a minus outside.)

An exact value for cos 15 degrees

Write 15 as a difference of angles we already know: 15 = 45 − 30.

1. Use the cosine-difference formula: cos(45 − 30) = cos45 cos30 + sin45 sin30. (Difference inside, so the outside sign is plus.)
2. Drop in exact values: (√2/2)(√3/2) + (√2/2)(1/2).
3. Multiply each piece: √6/4 + √2/4.
4. Add over the common bottom: (√6 + √2) ÷ 4.

A decimal check: (2.449 + 1.414) ÷ 4 ≈ 0.966, and a calculator agrees that cos 15° ≈ 0.966. The formula gave an exact answer no table could.

Double-angle formulas

Setting B equal to A in the sum formulas gives two more:

• sin 2θ = 2 sinθ cosθ.
• cos 2θ = cos²θ − sin²θ, which also equals 2cos²θ − 1 and 1 − 2sin²θ (three equal forms; use whichever fits).

Given sinθ = 3/5 in Quadrant I, so cosθ = 4/5:

1. sin 2θ = 2 × (3/5) × (4/5) = 24/25.
2. cos 2θ = (4/5)² − (3/5)² = 16/25 − 9/25 = 7/25.

Sanity check: (24/25)² + (7/25)² = (576 + 49) ÷ 625 = 1. The doubled angle still lands on the unit circle, as it must.

Where people get stuck

The single biggest trap is treating these functions as if they distribute, writing sin(A + B) = sinA + sinB. That is false. The sine of a sum needs the cross terms, sinA cosB + cosA sinB. Test it with numbers whenever you doubt it: sin(90°) = 1, but sin(45°) + sin(45°) = √2 ≈ 1.414, and those are not equal. The second trap is reading sin²θ as sin(θ²); it means the sine value squared. When a proof stalls, fall back on the safe habit: rewrite in sines and cosines and simplify.

Real-world reason this matters

Identities are quiet workhorses in physics and engineering. Adding two sound or radio waves uses the sum formulas; the double-angle form of a projectile's range, R = (v² sin 2θ) ÷ g, explains why 45 degrees throws a ball farthest; and power in alternating-current circuits is found by rewriting products of sinusoids with these same tools. Even your calculator leans on identity-based shortcuts to evaluate trig functions quickly.

Common misconceptions

• "sin(A + B) = sin A + sin B." False; the sum formula has the cross terms sinA cosB + cosA sinB.
• "sin²θ means sin(θ²)." It means (sinθ)², the square of the sine value.
• "You can cancel sine from both sides of an identity." Identities are proved by transforming one side, not by operating on both as if it were already known.
• "cos 2θ = 2 cosθ." Doubling the angle is not doubling the value; use one of the three double-angle forms.
• "There is one correct proof." Any valid chain that turns one side into the other is a complete proof.

Recap

The identity toolkit: reciprocal (sec = 1 ÷ cos), quotient (tan = sin ÷ cos), Pythagorean (sin² + cos² = 1 and its two divided forms), even-odd (cosine even, sine and tangent odd), sum and difference formulas for exact values like cos 15°, and double-angle formulas sin 2θ = 2 sinθ cosθ and cos 2θ in three forms. Prove identities by converting one side to sines and cosines and simplifying with these tools until it matches the other side. You now have every trade you need.

Sources

1. OpenStax. (2021). 7.1 Simplifying and verifying trigonometric identities. In Precalculus 2e. openstax.org
2. OpenStax. (2021). 7.2 Sum and difference identities. In Precalculus 2e. openstax.org
3. OpenStax. (2021). 7.3 Double-angle, half-angle, and reduction formulas. In Precalculus 2e. openstax.org
4. OpenStax. (2021). 7.4 Sum-to-product and product-to-sum formulas. In Precalculus 2e. openstax.org
5. OpenStax. (2021). 9.1 Verifying trigonometric identities. In Algebra and Trigonometry 2e. openstax.org
6. Khan Academy. (n.d.). Trigonometry. khanacademy.org
7. Math is Fun. (n.d.). Trigonometric identities. mathsisfun.com

If solving an equation with sine or cosine in it has ever felt like guessing in the dark, that is a completely normal place to start, and we are going to turn the lights on. Here is the encouraging truth: you already know how to solve these. They obey the same moves as ordinary algebra, isolate, factor, undo. The one new idea is that trig functions repeat, so a single equation can have several answers, sometimes infinitely many. We will handle that one gentle step at a time.

The big picture

A trig equation asks "which angles make this true?" Because sine and cosine cycle around and around, the same value turns up again and again, so we usually either look inside one trip around the circle or write a formula for all of them. This is exactly how you find when a tide, a daylight curve, or an electric current reaches a target value.

Key idea: solve it like algebra to get one or two angles in a single revolution, then repeat by the period to collect them all.

The recipe

1. Get the trig function alone, just like isolating a variable.
2. Find the reference angle from the value.
3. Place an answer in each quadrant where the function has the right sign.
4. If asked for every solution, add whole periods.

Let us solve one together, slowly

Solve 2 sinθ − 1 = 0 on 0 ≤ θ < 2π (read "theta from 0 up to, but not including, 2 pi").

1. Add 1 to both sides: 2 sinθ = 1. (Undo the minus 1.)
2. Divide both sides by 2: sinθ = 1/2. (Now the function stands alone.)
3. Recall the reference angle: sine equals 1/2 at π/6, which is 30 degrees.
4. Decide the quadrants: sine is positive in Quadrants I and II.
5. Write both answers: θ = π/6 and θ = π − π/6 = 5π/6.

Two answers, found the same careful way you would solve any equation. The only new part was step 4, thinking about which quadrants. Nice.

A cosine example

Solve cosθ = −1/2 on 0 ≤ θ < 2π.

1. Find the reference angle for cosine 1/2: it is π/3, or 60 degrees. (Use the size, set the sign aside for a moment.)
2. Cosine is negative in Quadrants II and III. (That is where the minus sign points us.)
3. Quadrant II answer: π − π/3 = 2π/3.
4. Quadrant III answer: π + π/3 = 4π/3.

Factor first when you can

Solve 2 sin²θ − sinθ = 0 on 0 ≤ θ < 2π.

1. Factor out the common sinθ: sinθ(2 sinθ − 1) = 0. (Same as pulling x out of an algebra expression.)
2. Use the zero-product rule: if a product is zero, one factor is zero. Set each to zero.
3. First factor: sinθ = 0 gives θ = 0 and θ = π.
4. Second factor: 2 sinθ − 1 = 0 gives sinθ = 1/2, so θ = π/6 and 5π/6.

All four values solve it. Notice the whole thing was algebra you already know, wearing a trig costume.

Try it: solve 2 cosθ = √3 on 0 ≤ θ < 2π.

Answer: Divide by 2: cosθ = √3/2, reference angle π/6. Cosine is positive in Quadrants I and IV, so θ = π/6 and 11π/6. Well done.

Writing every solution

To capture all the answers, add whole periods to each one. Sine and cosine repeat every 2π, so sinθ = 1/2 has the general solution θ = π/6 + 2πk or θ = 5π/6 + 2πk, where k is any whole number. Tangent repeats every π, so tanθ = 1 collapses into one tidy family, θ = π/4 + πk, which already covers both π/4 and 5π/4.

A quadratic in disguise

Solve 2cos²θ + cosθ − 1 = 0 on 0 ≤ θ < 2π. Treat cosθ like a single variable u, so it reads 2u² + u − 1 = 0.

1. Factor: (2u − 1)(u + 1) = 0, that is (2cosθ − 1)(cosθ + 1) = 0.
2. First factor: cosθ = 1/2, giving θ = π/3 and 5π/3.
3. Second factor: cosθ = −1, giving θ = π.

Three answers in one revolution: π/3, π, and 5π/3.

When the angle is doubled

Solve sin 2θ = 1 on 0 ≤ θ < 2π. Let u = 2θ. As θ runs from 0 to 2π, u runs from 0 to 4π, two full trips, so expect twice as many answers.

1. Solve for u first: sine equals 1 at u = π/2, and one period later at u = π/2 + 2π = 5π/2 (both under 4π).
2. Divide each by 2 to get θ: θ = π/4 and θ = 5π/4.

With a doubled angle, solve for the whole angle first and divide last. Doing it the other way around quietly loses answers.

Worked example: a sine quadratic, and why 2x doubles the count

Question: Solve 2sin²x − sin x − 1 = 0 on 0 ≤ x < 2π.

1. Substitute. Let u = sin x, so the equation reads 2u² − u − 1 = 0.
2. Factor. (2u + 1)(u − 1) = 0. Expand to check: 2u² − 2u + u − 1 = 2u² − u − 1. Correct.
3. First factor. sin x = −1/2. Sine is negative in Quadrants III and IV with reference angle π/6, giving x = 7π/6 and x = 11π/6.
4. Second factor. sin x = 1, which happens exactly once per revolution, at x = π/2.

Three solutions: π/2, 7π/6, and 11π/6. Notice that sin x = 1 gave only one answer while sin x = −1/2 gave two. Values at the very top or bottom of the wave are hit once per cycle; everything in between is hit twice.

Question: Solve cos 2x = 1/2 on 0 ≤ x < 2π.

1. Widen the window first. If x runs over [0, 2π), then 2x runs over [0, 4π) - two full revolutions.
2. Solve for 2x across that whole window. Cosine equals 1/2 at π/3 and 5π/3 in the first revolution, then again 2π later at 7π/3 and 11π/3. All four are under 4π.
3. Divide every answer by 2 last. x = π/6, 5π/6, 7π/6, 11π/6.

Check one. At x = 7π/6: 2x = 7π/3, which is coterminal with 7π/3 − 2π = π/3, and cos(π/3) = 1/2. It works.

Four solutions, not two. Whenever the angle is multiplied by k, expect roughly k times as many solutions in the same window, and always widen the window before you solve rather than trying to patch answers on afterwards.

When the value is not a special one

Solve sinθ = 0.4 on 0 ≤ θ < 2π. No special angle has sine 0.4, so ask the calculator: θ = sin⁻¹(0.4) ≈ 0.4115 radians (read "inverse sine of 0.4"). That is the Quadrant I answer. Sine is also positive in Quadrant II, so the second answer is π − 0.4115 ≈ 2.7301. The calculator hands you one angle; the quadrant thinking supplies the rest.

Where people get stuck

The most expensive mistake is dividing both sides by a trig factor. Solving sinθ cosθ = cosθ by dividing by cosθ looks tempting, but it silently throws away every angle where cosθ = 0. Instead, move everything to one side and factor: cosθ(sinθ − 1) = 0. Then cosθ = 0 gives π/2 and 3π/2, and sinθ = 1 gives π/2 (already on the list). The safe rule: never divide by something that might be zero, factor instead. The other common slip is stopping at one answer when the interval holds two.

Real-world reason this matters

Trig equations answer timing questions about anything that cycles: at what moments the tide is 2 meters above its average, on which day a city gets exactly 12 hours of daylight, at what angle a projectile reaches a target range, and when an alternating current crosses a safety threshold. In each case the model is a sinusoid, and the question "when does it equal this value?" is precisely a trig equation.

Common misconceptions

• "One answer is enough." Inside one revolution most values occur twice, once per quadrant with the right sign, and the general solution has infinitely many.
• "Divide both sides by a trig factor." Factoring is safe; dividing throws away the solutions where that factor is zero.
• "The calculator's inverse gives every solution." It gives one; use reference angles and quadrant signs to find the others.
• "For sin 2 theta, divide the interval answers by 2 at the start." Solve for the doubled angle across the stretched interval first, then divide.
• "Squaring both sides is harmless." Squaring can invent extra roots; check every candidate in the original equation.

Recap

Solve trig equations like algebra: isolate the trig function or factor (zero-product and quadratic-in-form patterns included), find the reference angle, place answers in the quadrants with the correct sign, and add the period for the general solution. For a doubled angle, solve for the whole angle across the stretched interval before dividing. Use inverse functions for non-special values, and never divide away a factor that could be zero. You have every tool this topic needs.

Sources

1. OpenStax. (2021). 7.5 Solving trigonometric equations. In Precalculus 2e. openstax.org
2. OpenStax. (2021). 7.1 Simplifying and verifying trigonometric identities. In Precalculus 2e. openstax.org
3. OpenStax. (2021). 9.5 Solving trigonometric equations. In Algebra and Trigonometry 2e. openstax.org
4. OpenStax. (2021). 6.3 Inverse trigonometric functions. In Precalculus 2e. openstax.org
5. OpenStax. (2021). 7.3 Double-angle, half-angle, and reduction formulas. In Precalculus 2e. openstax.org
6. Khan Academy. (n.d.). Trigonometry. khanacademy.org
7. Math is Fun. (n.d.). Trigonometric identities. mathsisfun.com

If solving a triangle sounds like something only surveyors can do, stay with me, because you are about to be able to do it too. Right triangles were friendly: SOH-CAH-TOA and you were done. But most triangles in the world are not right triangles, and for a long time that has felt like a dead end to a lot of students. It is not. Two clean formulas solve any triangle at all, and we will meet them one at a time, with every number worked out in the open.

The big picture

Two laws connect a triangle's three sides to its three angles. One is perfect when you already know an angle sitting across from a known side; the other is perfect when you do not. Between them, if a triangle can be solved, these two will solve it. That is how you measure a distance you cannot walk across, like the width of a river.

Key idea: use the law of sines when you have a matching angle-and-opposite-side pair; otherwise reach for the law of cosines.

Labeling a triangle

By custom the three sides are lowercase a, b, c, and each faces the angle with the matching capital letter: side a is across from angle A, side b from angle B, side c from angle C. Keeping this pairing straight is half the battle, so it is worth a slow look before any arithmetic.

The law of sines

The law of sines says each side is in step with the sine of its opposite angle:

sin A ÷ a = sin B ÷ b = sin C ÷ c

Use it when you know an angle and the side across from it, plus one more piece. That covers angle-angle-side (AAS), angle-side-angle (ASA), and the tricky side-side-angle (SSA).

Let us solve one together, slowly

A triangle has A = 40°, B = 60°, and side a = 10. Find side b.

1. Set up the matching pairs: sin 40° ÷ 10 = sin 60° ÷ b. (Each side over the sine of its opposite angle.)
2. Cross multiply: b × sin 40° = 10 × sin 60°. (Clear the fractions.)
3. Divide to get b alone: b = (10 × sin 60°) ÷ sin 40°.
4. Put in decimals: b ≈ (10 × 0.866) ÷ 0.643 ≈ 13.5.

So side b is about 13.5. Every step was a move you already know from solving proportions.

The law of cosines

When there is no angle-opposite-side pair to start from, the law of cosines steps in. It is the Pythagorean theorem with a correction term for triangles that are not right-angled:

c² = a² + b² − 2ab · cos C

Use it for side-angle-side (SAS), two sides and the angle wedged between them, or side-side-side (SSS), all three sides when you want an angle.

A side-angle-side example

Two sides are a = 5 and b = 7 with the angle between them C = 60°. Find side c.

1. Write the law: c² = 5² + 7² − 2(5)(7) cos 60°.
2. Square and multiply: c² = 25 + 49 − 70 × 0.5.
3. Simplify: c² = 74 − 35 = 39.
4. Take the square root: c = √39 ≈ 6.2.

A reassuring check: if C were 90°, then cos 90° = 0 and the formula would shrink to c² = a² + b², the plain Pythagorean theorem. So this really is Pythagoras with an adjustment.

Try it: two sides 8 and 6 meet at a 90° angle. Use the law of cosines to find the third side.

Answer: c² = 8² + 6² − 2(8)(6) cos 90° = 64 + 36 − 0 = 100, so c = 10, matching the Pythagorean theorem exactly. Well done.

Choosing the right law

You are given	Use
AAS or ASA (an angle-side opposite pair)	Law of sines
SSA (ambiguous case)	Law of sines, then check for two triangles
SAS (two sides and the included angle)	Law of cosines
SSS (three sides)	Law of cosines

Finding an angle from three sides

A triangle has a = 7, b = 8, c = 9. Find angle C, the one across from the longest side.

1. Rearrange the law of cosines to solve for the cosine: cos C = (a² + b² − c²) ÷ (2ab).
2. Substitute: cos C = (49 + 64 − 81) ÷ (2 × 7 × 8).
3. Simplify: cos C = 32 ÷ 112 = 2/7 ≈ 0.2857.
4. Undo the cosine: C = cos⁻¹(2/7) ≈ 73.4°.

A quick sanity check: the longest side (9) should face the biggest angle, and 73.4° is indeed the largest here. A nice bonus of solving for cosine, a negative result would announce an obtuse angle right away, with no quadrant guessing.

The ambiguous case (when two triangles fit)

Given A = 30°, a = 5, b = 8 (an SSA setup), find angle B.

1. Law of sines: sin B = (b × sin A) ÷ a = (8 × 0.5) ÷ 5 = 0.8.
2. Two different angles between 0 and 180 degrees have sine 0.8, so we test both.
3. First: B ≈ 53.1°, leaving C = 180° − 30° − 53.1° = 96.9°. Valid.
4. Second: B ≈ 180° − 53.1° = 126.9°, leaving C = 180° − 30° − 126.9° = 23.1°. Also valid.

Both leave a positive third angle, so two whole triangles fit the very same numbers. In any SSA problem, always test the supplement: if it still leaves a positive third angle, a second triangle exists, and if sin B ever comes out above 1, no triangle exists at all.

Worked example: the ambiguous case with real numbers

"Two triangles fit" sounds abstract until you build both. Question: a = 8, b = 10, and angle A = 40 degrees. Solve every triangle that fits.

1. Find sin B by the law of sines. sin B / b = sin A / a, so sin B = (10 sin 40°)/8 = (10)(0.6428)/8 = 0.8035.
2. Two angles share that sine. The calculator gives B ≈ 53.5°, and its supplement 180 - 53.5 = 126.5° has the same sine.
3. Test each against the 180-degree budget. 40 + 53.5 = 93.5, leaving 86.5 degrees - fine. 40 + 126.5 = 166.5, leaving 13.5 degrees - also fine. Both work, so there really are two triangles.
4. Finish triangle 1. C = 86.5°, and c = a sin C / sin A = 8(0.9981)/0.6428 ≈ 12.4.
5. Finish triangle 2. C = 13.5°, and c = 8(0.2334)/0.6428 ≈ 2.9.

One data set, two genuinely different triangles: a long thin one with c about 12.4 and a squat one with c about 2.9. Both satisfy every given measurement.

When does this happen? Only in the side-side-angle case, where the given angle is not between the two given sides. And only sometimes even then: if the supplement pushes the angle total past 180 degrees, the second triangle is impossible and you discard it. The test in step 3 is not optional bookkeeping - it is the entire ambiguous case.

Where people get stuck: accepting whatever the calculator returns. sin⁻¹ always hands back an angle between 0 and 90 degrees for a positive input, so it can only ever show you the acute possibility. The obtuse one exists whenever 180 − B still fits, and only you can check that. The law of cosines never has this problem, because cos⁻¹ spans the full 0 to 180 range and returns a unique angle - which is a good reason to prefer it whenever you have a choice.

The area shortcut

Two sides and the angle between them also give the area straight away: Area = ½ · a · b · sin C. For sides 5 and 7 meeting at 60°: Area = ½ × 5 × 7 × sin 60° ≈ 17.5 × 0.866 ≈ 15.2 square units. When C is 90°, sin 90° = 1 and this becomes the familiar half base times height.

Where people get stuck

Two traps catch most people. First, trying to start with the law of sines when there is no angle-opposite-side pair; without that anchor the proportion has two unknowns and stalls, so begin with the law of cosines instead. Second, forgetting that the calculator's inverse sine returns only one angle. In an SSA problem the supplement may give a second valid triangle, so always check it. One steady habit prevents both: match each side to its opposite angle first, and remember the longest side always faces the largest angle.

Real-world reason this matters

These two laws are the mathematics of measuring things you cannot reach. Surveyors find the width of a lake by measuring two distances and the angle between them from a single shore station (SAS, law of cosines). Sailors and pilots fix their position from two bearings to known landmarks (ASA, law of sines). Cell networks locate phones, astronomers gauge distances to nearby stars by parallax, and engineers resolve the forces in a bridge truss, all with these same equations.

Common misconceptions

• "The law of sines works with any three pieces of data." It needs a known angle-opposite-side pair to anchor the proportion; otherwise start with the law of cosines.
• "The calculator's inverse-sine answer is the only angle." Sine is positive in two quadrants, so SSA problems must also test the supplement.
• "The law of cosines needs a right angle." It works in every triangle; the right-angle case simply reduces it to the Pythagorean theorem.
• "Any angle can pair with any two sides in SAS." The angle must be the included angle, sitting between the two given sides.
• "Sides and angles can be matched however you like." Each side faces its own opposite angle, and the largest side always faces the largest angle; use this to catch errors.

Recap

The law of sines (sin A ÷ a = sin B ÷ b = sin C ÷ c) solves triangles with a known angle-opposite-side pair (AAS, ASA, and the ambiguous SSA, which may give two triangles). The law of cosines (c² = a² + b² − 2ab cos C) handles SAS and SSS, and its rearranged form finds an angle with no ambiguity. The area of any SAS triangle is ½ ab sin C. Together the two laws solve every solvable triangle, and now they are yours.

Sources

1. OpenStax. (2021). 8.1 Non-right triangles: Law of sines. In Precalculus 2e. openstax.org
2. OpenStax. (2021). 8.2 Non-right triangles: Law of cosines. In Precalculus 2e. openstax.org
3. OpenStax. (2021). 10.1 Non-right triangles: Law of sines. In Algebra and Trigonometry 2e. openstax.org
4. OpenStax. (2021). 10.2 Non-right triangles: Law of cosines. In Algebra and Trigonometry 2e. openstax.org
5. OpenStax. (2021). 6.3 Inverse trigonometric functions. In Precalculus 2e. openstax.org
6. Khan Academy. (n.d.). Trigonometry. khanacademy.org
7. Math is Fun. (n.d.). Solving triangles. mathsisfun.com

If the word "sequence" sounds fancier than it needs to, let us take the mystery out of it right away. A sequence is just a list of numbers with a pattern, the kind you have followed since you were small: 2, 4, 6, 8. If you can count by twos, you already understand the heart of this lesson. We are only going to name a couple of patterns you already recognize, and learn a shortcut for jumping far ahead in the list without writing out every number.

The big picture

Two patterns show up everywhere, so they earn names. One adds the same amount each step; the other multiplies by the same amount each step. Once you can tell which is which, a single formula finds the 100th term without writing the first 99. That is the whole payoff.

Key idea: add the same number each time and it is arithmetic; multiply by the same number each time and it is geometric.

Sequences, in plain words

A sequence is an ordered list of numbers, each one called a term. We label them a₁, a₂, a₃, read "a-one, a-two, a-three," so a₁ is the first term, a₂ the second, and so on. The little numbers are position labels, like seat numbers in a row.

Arithmetic: adding the same step

An arithmetic sequence adds a fixed number, the common difference d, to move from one term to the next. In 3, 7, 11, 15 you add 4 each time, so d = 4. To jump to any term without listing them all, use:

aₙ = a₁ + (n − 1)d

Read that as "the nth term equals the first term plus (n minus 1) steps of size d." The n − 1 is there because the first term has taken zero steps yet.

Let us find a term together, slowly

Find the 10th term of 3, 7, 11, 15, ...

1. Spot the first term: a₁ = 3. (The list starts at 3.)
2. Spot the common difference: each step adds 4, so d = 4. (7 minus 3 is 4.)
3. Count the steps to the 10th term: n − 1 = 10 − 1 = 9. (Nine hops from the first term.)
4. Put it together: a₁₀ = 3 + 9 × 4 = 3 + 36 = 39.

So the 10th term is 39, found without writing out all ten. That formula just saved you a lot of counting.

Geometric: multiplying the same factor

A geometric sequence multiplies by a fixed number, the common ratio r, to advance. In 2, 6, 18, 54 you multiply by 3 each time, so r = 3. Its jump-ahead formula is:

aₙ = a₁ · rⁿ⁻¹

Read the exponent n − 1 the same way, as the number of times you multiply.

Find the 5th term of 2, 6, 18, 54, ...

1. First term: a₁ = 2.
2. Common ratio: r = 3 (6 divided by 2 is 3).
3. Number of multiplications: n − 1 = 4.
4. Compute: a₅ = 2 × 3⁴ = 2 × 81 = 162.

Telling the two apart

To classify a list, look at what happens between neighbors. If the difference is always the same, it is arithmetic. If the ratio is always the same, it is geometric. So 5, 10, 20, 40 is geometric (times 2), while 5, 10, 15, 20 is arithmetic (plus 5). Some lists are neither, like the squares 1, 4, 9, 16, whose gaps 3, 5, 7 keep changing.

Try it: find the 8th term of the geometric sequence 3, 6, 12, 24, ...

Answer: a₁ = 3, r = 2, and n − 1 = 7, so a₈ = 3 × 2⁷ = 3 × 128 = 384. Well done.

Recursive rules (one term from the last)

Besides the jump-ahead formula, a sequence can be defined recursively: state the first term, then a rule that builds each term from the one before. Arithmetic example: a₁ = 2 and aₙ = aₙ₋₁ + 7 makes 2, 9, 16, 23, ... Geometric example: a₁ = 5 and aₙ = 2 · aₙ₋₁ makes 5, 10, 20, 40, ... Recursive rules match how a spreadsheet fills a column one cell at a time, while the jump-ahead formula leaps straight to term 1000 without the first 999.

Working backward from two terms

An arithmetic sequence has a₃ = 11 and a₇ = 23. Find d and a₁.

1. Count the steps between them: term 3 to term 7 is 4 steps. (7 minus 3.)
2. Those 4 steps cover 23 − 11 = 12, so 4d = 12. (Total change across the steps.)
3. Divide: d = 3.
4. Back up from a₃: a₁ = 11 − 2 × 3 = 5. (Two steps back from term 3.)

Check: a₇ = 5 + 6 × 3 = 23. It matches, so we can trust it. For a geometric one with a₂ = 6 and a₅ = 48: term 2 to term 5 is 3 multiplications, so r³ = 48 ÷ 6 = 8, giving r = 2. Then a₁ = 6 ÷ 2 = 3, and the list runs 3, 6, 12, 24, 48.

Where the two explicit formulas come from

Both closed-form rules are just the recursive rule applied over and over, written down.

Arithmetic. Start at a₁ and add d each step:

• a₂ = a₁ + d (one d added)
• a₃ = a₁ + d + d = a₁ + 2d (two d's)
• a₄ = a₁ + 3d (three d's)

The pattern is unmissable: to reach term n you take n − 1 steps, so aₙ = a₁ + (n − 1)d. That n − 1 is not a quirk to memorize; it counts gaps, and there is always one fewer gap than terms - the same reason a fence with 10 posts has 9 sections.

Geometric. Same reasoning, with multiplying instead of adding: a₂ = a₁r, a₃ = a₁r², a₄ = a₁r³, so aₙ = a₁ · rn-1. Again the exponent counts the multiplications, and there are n − 1 of them.

Worked check. The sequence 5, 8, 11, 14, ... has a₁ = 5 and d = 3. The formula predicts a₄ = 5 + 3(3) = 14, which is the fourth listed term. And a₁₀ = 5 + 9(3) = 32. Counting up by 3 from 14 confirms: 17, 20, 23, 26, 29, 32. It lands.

Which term hits a target?

In 3, 7, 11, ..., which term equals 99? Set the formula equal to 99 and solve for n.

1. 3 + (n − 1) × 4 = 99. (The nth term is 99.)
2. Subtract 3: (n − 1) × 4 = 96.
3. Divide by 4: n − 1 = 24.
4. Add 1: n = 25. So 99 is the 25th term.

If solving ever gives a fraction for n, that just means the target is not on the list, which is a perfectly good answer too.

Ratios can be negative or fractional

The common ratio does not have to be a whole number or positive. The sequence 16, −8, 4, −2, ... has r = −1/2: each term is half the size and flips its sign. Alternating plus and minus signs are the fingerprint of a negative ratio.

Where people get stuck

The number one slip is using n instead of n − 1 in the formulas. It is tempting to multiply d by n, but the first term has taken zero steps, so the count is always one less than the term number. A quick self-check: the formula should hand back a₁ when n = 1, and a₁ + (1 − 1)d = a₁ confirms it. The other common mix-up is assuming any growing list is geometric. Arithmetic lists grow too, so always test whether the difference or the ratio is the steady one.

Real-world reason this matters

Arithmetic sequences model steady, repeated addition: a salary that starts at 40,000 dollars with a 2,000 dollar raise each year, the seats added row by row in a theater, fence posts along a road, and simple interest. Geometric sequences model proportional change: compound interest, a population that grows by a fixed percent, radioactive material halving over time, and the doubling thickness of folded paper. The first question in any of these is always the same, is the same amount being added, or the same factor multiplied?

Common misconceptions

• "The nth-term formula uses n, so multiply d by n." The factor is (n - 1): the first term has taken zero steps.
• "Ratios must be bigger than 1." Ratios can be fractions (shrinking) or negative (alternating signs).
• "If it grows, it is geometric." Arithmetic sequences grow too; test whether the difference or the ratio is constant.
• "Recursive and explicit formulas are rivals." They describe the same sequence; recursive is easy to state, the jump-ahead form is fast at large n.
• "A sequence must start at a small number." Any first term works; the pattern, not the starting value, sets the type.

Recap

A sequence is an ordered list. Arithmetic sequences add a common difference d, with aₙ = a₁ + (n − 1)d; geometric sequences multiply by a common ratio r, with aₙ = a₁ · rⁿ⁻¹. Recursive rules build each term from its predecessor. Two known terms recover d (divide the change by the number of steps) or r (take the matching root of the quotient), and setting the formula equal to a target value tells you which term it is. You now read both patterns with confidence.

Sources

1. OpenStax. (2021). 11.2 Arithmetic sequences. In Precalculus 2e. openstax.org
2. OpenStax. (2021). 11.3 Geometric sequences. In Precalculus 2e. openstax.org
3. OpenStax. (2021). 11.1 Sequences and their notations. In Precalculus 2e. openstax.org
4. OpenStax. (2021). 13.2 Arithmetic sequences. In Algebra and Trigonometry 2e. openstax.org
5. OpenStax. (2021). 13.3 Geometric sequences. In Algebra and Trigonometry 2e. openstax.org
6. Khan Academy. (n.d.). Precalculus. khanacademy.org
7. Math is Fun. (n.d.). Sequences and series. mathsisfun.com

Adding up a long list of numbers sounds tedious, but we are going to learn shortcuts that do the whole thing in one line. A series is just the total when you add up a sequence, and you have added lists your whole life. The only new part is a compact way to write "add these up" and two tidy formulas that spare you from adding term by term.

The big picture

A sequence is a list; a series is its running total. Instead of adding a hundred numbers by hand, a couple of formulas give the sum in one step, as long as the list is arithmetic or geometric. That is genuinely useful for loans, savings, and any repeated payment.

Key idea: a series is the sum of a sequence, and if the sequence is arithmetic or geometric, a single formula gives the total.

Sigma notation (a compact "add these up")

Mathematicians write sums with the Greek capital sigma, which just means "add up." A sum from k = 1 to 4 of k tells you to plug in k = 1, 2, 3, 4 and add: 1 + 2 + 3 + 4 = 10. The letter k is only a counter; the bottom and top numbers say where to start and where to stop. Do not let the symbol worry you, it is shorthand for something you already do.

Adding an arithmetic series

There is a lovely shortcut. Pair the first term with the last, the second with the second-to-last, and each pair has the same total. That gives:

Sₙ = (n ÷ 2)(a₁ + aₙ)

In words: the number of terms times the average of the first and last term.

Let us add one together, slowly

Add all the whole numbers from 1 to 100.

1. Identify the pieces: a₁ = 1, the last term aₙ = 100, and there are n = 100 terms.
2. Add first and last: 1 + 100 = 101. (Every pair totals this.)
3. Multiply by half the count: S = (100 ÷ 2) × 101 = 50 × 101.
4. Finish: 50 × 101 = 5050.

That is the sum of a hundred numbers in four small steps. Legend says a schoolboy named Gauss found this very trick and finished the task in seconds.

Where the arithmetic sum formula comes from

The formula is not something to accept on faith. There is a story about the young Carl Friedrich Gauss being told to add 1 through 100 and answering in seconds, and whether or not the story is exact, the trick is real.

Write the sum forwards and then backwards, one line under the other:

S = 1 + 2 + 3 + ... + 98 + 99 + 100

S = 100 + 99 + 98 + ... + 3 + 2 + 1

Now add the two lines column by column. Every column totals 101: 1 + 100, 2 + 99, 3 + 98, all the way down. There are 100 columns, so

2S = 100 x 101 = 10 100, so S = 5050

Nothing about that argument used the numbers 1 and 100 specifically. Run it on any arithmetic series with n terms, first term a₁ and last term aₙ: every column sums to a₁ + aₙ, and there are n of them, so 2S = n(a₁ + aₙ), giving

S = n(a₁ + aₙ) / 2

Read the formula in English and it is obvious: the number of terms, times the average of the first and last. That is why an arithmetic series is so easy to add - the terms are spread evenly, so the average of the whole list equals the average of just its two ends.

Where the geometric sum formula comes from

Even sums cannot be paired that way, because geometric terms are not evenly spaced. A different trick works: multiply by r and subtract.

1. Write the sum: S = a + ar + ar² + ... + arn-1.
2. Multiply every term by r: rS = ar + ar² + ... + arn-1 + arn.
3. Subtract the second line from the first. Almost everything cancels in the middle, leaving only the first term of one and the last term of the other: S − rS = a − arn.
4. Factor both sides: S(1 − r) = a(1 − rn).
5. Divide: S = a(1 − rn) / (1 − r), valid whenever r ≠ 1.

Test it on something you can check by hand. Sum 2 + 6 + 18 + 54, where a = 2, r = 3, n = 4. The formula gives 2(1 − 81)/(1 − 3) = 2(-80)/(-2) = 80. Adding directly: 2 + 6 + 18 + 54 = 80. It matches.

And notice what happens as n grows when |r| < 1: the term rn shrinks toward zero, so the whole formula settles to S = a/(1 − r). That is exactly the infinite-series formula in the next section, and now you can see where it comes from rather than memorizing it.

Adding a geometric series

For a geometric list (with r ≠ 1), the sum of the first n terms is:

Sₙ = a₁ · (1 − rⁿ) ÷ (1 − r)

Add 2 + 6 + 18 + 54 + 162, where a₁ = 2, r = 3, n = 5:

1. Compute rⁿ = 3⁵ = 243. (The ratio to the 5th power.)
2. Top: 1 − 243 = −242.
3. Bottom: 1 − 3 = −2.
4. Put it together: S = 2 × (−242) ÷ (−2) = 2 × 121 = 242.

Adding the five numbers by hand also gives 242, so the formula checks out.

Infinite geometric series (a surprising, finite total)

Here is something that feels impossible at first: you can add infinitely many numbers and still get a finite answer, as long as the terms shrink fast enough. If |r| < 1 (the ratio's size is less than 1), the series converges to:

S = a₁ ÷ (1 − r)

For 1 + 1/2 + 1/4 + 1/8 + ...: here a₁ = 1 and r = 1/2, so S = 1 ÷ (1 − 1/2) = 1 ÷ (1/2) = 2. If instead |r| ≥ 1, the terms never shrink and the series diverges, with no finite total.

Watching it converge

The partial sums of 1 + 1/2 + 1/4 + ... creep toward 2 and never pass it:

Terms added	1	2	3	4	5
Partial sum	1	1.5	1.75	1.875	1.9375

Each new term closes half the remaining gap to 2. Saying "the series sums to 2" means exactly that these running totals head toward 2.

Try it: find the sum of the infinite geometric series 6 + 2 + 2/3 + ...

Answer: a₁ = 6 and r = 1/3 (each term is a third of the last), so S = 6 ÷ (1 − 1/3) = 6 ÷ (2/3) = 9. Well done.

Reading sigma notation carefully

Evaluate the sum from k = 1 to 5 of (2k + 1). Substitute each k and add: 3 + 5 + 7 + 9 + 11 = 35. The counter can be named k or i or j; it changes nothing. The start and stop numbers are what matter.

A repeating decimal is secretly a series

Write 0.7777... as a fraction.

1. Unpack the decimal: 7/10 + 7/100 + 7/1000 + .... (Each 7 sits in the next place.)
2. Spot the pattern: geometric with a₁ = 7/10 and r = 1/10.
3. Use the infinite-sum formula: S = (7/10) ÷ (1 − 1/10) = (7/10) ÷ (9/10).
4. Simplify: S = 7/9.

Check by long division: 7 divided by 9 really is 0.777..., so a repeating decimal is just a geometric series in disguise.

A bouncing ball

A ball drops 10 m and rebounds to 60 percent of each previous height, forever. What total distance does it travel?

1. The rebound heights form a geometric list: 6, 3.6, 2.16, ... with r = 0.6.
2. Their sum is 6 ÷ (1 − 0.6) = 6 ÷ 0.4 = 15 m.
3. Each rebound is traveled twice, up and back down: 2 × 15 = 30 m.
4. Add the first 10 m drop: 10 + 30 = 40 m total.

Infinitely many bounces, yet a finite 40 meters. That is convergence you can actually see.

Where people get stuck

The first common error is thinking infinitely many terms must add to infinity. If the ratio's size is under 1, the pieces shrink fast enough for a finite total. The second is using the infinite-sum formula when |r| ≥ 1; then the series diverges and the formula is meaningless, so always check the ratio first. A smaller but frequent slip in the arithmetic formula is putting some middle term where aₙ belongs; that second slot must be the actual last term you are adding.

Real-world reason this matters

Finite geometric sums price loans and annuities, because each payment's value forms a geometric list. Arithmetic sums count stacked logs, theater seats, and total pay under steady raises. Infinite geometric series turn repeating decimals into fractions, add up repeated medicine doses approaching a steady level, and even model the economic multiplier, where one dollar spent circulates as a shrinking geometric series. Convergence, adding forever yet landing on a number, is also your first real glimpse of the limits that open calculus.

Common misconceptions

• "Infinitely many terms means an infinite sum." If |r| < 1 the pieces shrink fast enough for a finite total.
• "The infinite-sum formula always applies." It requires |r| < 1; for r = 3 or r = -2 the series diverges and the formula means nothing.
• "(n/2)(a₁ + aₙ) works with any term at the end." The second entry must be the nth term, the last one actually added.
• "A sequence and a series are the same thing." A sequence is the list; a series is the sum of the list.
• "Partial sums equal the limit." They approach it; each partial sum of a positive series still falls a little short of the total.

Recap

Sigma notation compresses a sum: read the start and stop, substitute each counter value, and add. Arithmetic series sum by pairing: S = (n ÷ 2)(first + last). Finite geometric series use S = a₁(1 − rⁿ) ÷ (1 − r), and infinite ones converge exactly when |r| < 1, to S = a₁ ÷ (1 − r). Repeating decimals, bouncing balls, and annuities are geometric series in costume, and the way partial sums settle toward a total is a first taste of limits.

Sources

1. OpenStax. (2021). 11.4 Series and their notations. In Precalculus 2e. openstax.org
2. OpenStax. (2021). 13.4 Series and their notations. In Algebra and Trigonometry 2e. openstax.org
3. OpenStax. (2021). 11.2 Arithmetic sequences. In Precalculus 2e. openstax.org
4. OpenStax. (2021). 11.3 Geometric sequences. In Precalculus 2e. openstax.org
5. Khan Academy. (n.d.). Precalculus. khanacademy.org
6. Dawkins, P. (n.d.). Series: The basics. Paul's Online Math Notes. Lamar University. tutorial.math.lamar.edu
7. Math is Fun. (n.d.). Geometric sequences and sums. mathsisfun.com

If vectors sound like something from a physics class you were dreading, let us make friends with them slowly. A vector is really just an arrow: it tells you how far and which way. Think of giving directions as "three blocks east and four blocks north." That pair of instructions, a distance together with a direction, is a vector. You have used them every time you pointed someone down the street. We will build up the arithmetic one gentle piece at a time.

The big picture

A plain number tells you only "how much." A vector tells you "how much and which way." Writing it as two components, a sideways part and an up-down part, lets you do geometry with simple arithmetic: add arrows by adding their parts. That is why vectors run navigation, physics, and computer graphics.

Key idea: a vector is an arrow stored as two numbers, ⟨sideways, up⟩, and you handle it one component at a time.

What a vector is

A vector has both a size (its length) and a direction, unlike a plain number, which has size only. In the plane we write it by its components, the horizontal amount and the vertical amount, as v = ⟨a, b⟩, read "the vector a, b." Picture an arrow starting at the origin and ending at the point (a, b).

Length (magnitude)

The magnitude is the arrow's length, and it comes straight from the Pythagorean theorem, since the components form a right triangle:

|v| = √(a² + b²)

For v = ⟨3, 4⟩: |v| = √(9 + 16) = √25 = 5. The bars mean "length of," read "the magnitude of v."

Adding, subtracting, and scaling

Every vector operation works one component at a time, which is what makes them friendly.

• Add: ⟨a, b⟩ + ⟨c, d⟩ = ⟨a + c, b + d⟩. (Sideways with sideways, up with up.)
• Subtract: ⟨a, b⟩ − ⟨c, d⟩ = ⟨a − c, b − d⟩.
• Scale: k⟨a, b⟩ = ⟨ka, kb⟩, which stretches the arrow, or flips it if k is negative.

Let us combine two together, slowly

Let u = ⟨2, −1⟩ and v = ⟨3, 5⟩. Find 2u − v.

1. Scale u by 2: 2u = ⟨2 × 2, 2 × (−1)⟩ = ⟨4, −2⟩. (Multiply each part by 2.)
2. Line up the subtraction: ⟨4, −2⟩ − ⟨3, 5⟩.
3. Subtract the sideways parts: 4 − 3 = 1.
4. Subtract the up parts: −2 − 5 = −7.

So 2u − v = ⟨1, −7⟩. Notice we never mixed the two columns; that separation is the whole trick.

The dot product

The dot product of two vectors is a single number, found by multiplying matching parts and adding: u · v = a₁a₂ + b₁b₂. It measures how much two arrows point the same way, and it links to the angle θ between them by u · v = |u| |v| cosθ. The most useful consequence: two arrows are perpendicular exactly when their dot product is 0, because cos 90° = 0.

Dot product and angle, step by step

Let u = ⟨1, 2⟩ and v = ⟨3, −1⟩. Find the angle between them.

1. Dot product: u · v = (1)(3) + (2)(−1) = 3 − 2 = 1. (Multiply matching parts, then add.)
2. Lengths: |u| = √5 and |v| = √10.
3. Divide to get the cosine: cosθ = 1 ÷ (√5 × √10) = 1 ÷ √50 ≈ 0.141.
4. Undo the cosine: θ = cos⁻¹(0.141) ≈ 81.9°.

A small positive dot product means the arrows lean the same general way but sit nearly at a right angle, which the 81.9° confirms.

Try it: for u = ⟨6, 8⟩ and v = ⟨−8, 6⟩, find |u| and the dot product, then say whether they are perpendicular.

Answer: |u| = √(36 + 64) = √100 = 10. Dot product = (6)(−8) + (8)(6) = −48 + 48 = 0, so yes, they are perpendicular. Well done.

Worked example: from magnitude and direction to components

Problems usually describe a vector the way a person would - "50 newtons at 30 degrees above the horizontal" - but the arithmetic needs components. Trigonometry translates.

For a vector of magnitude r at angle θ from the positive x-axis:

v = < r cosθ, r sinθ >

Question: A rope pulls a crate with a force of 50 N at 30 degrees above the horizontal. Find the horizontal and vertical components.

1. Horizontal: 50 cos 30° = 50(0.8660) ≈ 43.3 N.
2. Vertical: 50 sin 30° = 50(0.5) = 25.0 N.
3. So F = <43.3, 25.0>.

Check by going back. The magnitude should return to 50: √(43.3² + 25.0²) = √(1875 + 625) = √2500 = 50. Exactly right.

Read the physics in those numbers. Only 43.3 N of the 50 actually drags the crate forward; the other 25 N lifts it. Pull at a lower angle and more of your effort goes into moving it. That trade-off is the whole reason engineers decompose forces into components.

Worked example: the reverse trip

Question: Convert v = <-6, 8> into magnitude and direction.

1. Magnitude. |v| = √((-6)² + 8²) = √(36 + 64) = √100 = 10.
2. Reference angle. tan⁻¹(8/6) = tan⁻¹(1.333) ≈ 53.1°.
3. Fix the quadrant. The components are left and up, so the vector is in Quadrant II. The calculator's answer must be measured from the negative x-axis, giving 180 − 53.1 = 126.9°.

Check. 10 cos 126.9° = 10(-0.6) = -6 and 10 sin 126.9° = 10(0.8) = 8. Both components come back, so the direction is right.

Step 3 is the one people skip. tan⁻¹ only ever returns angles between -90 and 90 degrees, so it cannot distinguish <-6, 8> from <6, -8> - both have the same ratio. Always sketch the components first and let the picture decide the quadrant.

Unit vectors and i, j

A unit vector has length 1 and records pure direction. To get the unit vector pointing the same way as v, divide v by its own length. For v = ⟨3, 4⟩ with length 5, the unit vector is ⟨3/5, 4/5⟩, and a quick check gives √(9/25 + 16/25) = √1 = 1. Two special unit vectors get names: i = ⟨1, 0⟩ (one step east) and j = ⟨0, 1⟩ (one step north), so any vector can also be written ⟨3, 4⟩ = 3i + 4j.

Vectors between points

The vector from point P(1, 2) to point Q(4, 6) is found by subtracting coordinates, head minus tail: ⟨4 − 1, 6 − 2⟩ = ⟨3, 4⟩, with length 5. Points name places; vectors name trips. The trip from (0, 0) to (3, 4) and the trip from (10, 10) to (13, 14) are the same vector, because the displacement is identical.

Direction angles

A vector's direction angle θ is measured counterclockwise from the positive x-axis, found from tanθ = b ÷ a with a check for the quadrant. For ⟨3, 3⟩: tanθ = 1 in Quadrant I, so θ = 45°, and the length is √(9 + 9) = 3√2. Going the other way, a vector of length m at angle θ has components ⟨m cosθ, m sinθ⟩.

An airplane in wind

A plane flies east at 200 mph while the wind blows north at 50 mph. Its true motion over the ground is the sum of the two arrows.

1. Add the vectors: ⟨200, 0⟩ + ⟨0, 50⟩ = ⟨200, 50⟩. (East part plus north part.)
2. Ground speed is the length: √(200² + 50²) = √(40000 + 2500) = √42500 ≈ 206.2 mph.
3. Direction: tanθ = 50 ÷ 200 = 0.25, so θ ≈ 14.0° north of east.

The plane ends up a little faster than its own airspeed and nudged off its heading, which is why pilots plan a correction. That correction is just one more vector added in.

Work as a dot product

In physics, the work done by a force F pushing something along a displacement d is W = F · d. If F = ⟨10, 5⟩ newtons and d = ⟨4, 0⟩ meters, then W = (10)(4) + (5)(0) = 40 joules. Only the part of the force lined up with the motion does work, and the dot product measures exactly that part.

Where people get stuck

The most common mix-up is expecting the dot product to be a vector. It is not; it is a single number, so if your answer has components, something slipped. The second frequent trap is assuming |u + v| = |u| + |v|. That is only true when the arrows point the exact same way; in general the combined arrow is shorter, the same reason two people pulling at an angle move a load less than their strengths simply added. When finding a direction angle, also watch the quadrant, because the calculator's inverse tangent needs an adjustment when the sideways component is negative.

Real-world reason this matters

Vectors carry every quantity that has both a size and a direction: the velocity of a plane or boat in a current, the forces in a tug-of-war or a bridge, the displacement in a navigation problem, even the motion of a character in a video game. Dot products test whether two directions are perpendicular in graphics engines, compute how light strikes a surface, and, as we just saw, define mechanical work.

Common misconceptions

• "A vector is a point." A vector is a displacement; it has no fixed location and can be drawn anywhere.
• "|u + v| = |u| + |v|." Only when the arrows point the same way; in general the sum is shorter.
• "The dot product is a vector." It is a single number; do not expect components.
• "Inverse tangent gives the direction angle directly." It needs a quadrant adjustment when the sideways component is negative.
• "Scaling changes direction." A positive scalar rescales length only; a negative scalar also flips the arrow around.

Recap

A vector ⟨a, b⟩ stores a length √(a² + b²) and a direction (from tanθ = b ÷ a with quadrant care). Add, subtract, and scale one component at a time, build a vector between points by head minus tail, and normalize to a unit vector by dividing by the length. The dot product a₁a₂ + b₁b₂ equals |u||v|cosθ, signals perpendicular at zero, and computes work in physics. Adding vectors turns navigation and force problems into plain arithmetic, and now that toolkit is yours.

Sources

1. OpenStax. (2021). 8.8 Vectors. In Precalculus 2e. openstax.org
2. OpenStax. (2021). 10.8 Vectors. In Algebra and Trigonometry 2e. openstax.org
3. OpenStax. (2021). 5.4 Right triangle trigonometry. In Precalculus 2e. openstax.org
4. Dawkins, P. (n.d.). Dot product. Paul's Online Math Notes. Lamar University. tutorial.math.lamar.edu
5. Dawkins, P. (n.d.). Vectors: The basics. Paul's Online Math Notes. Lamar University. tutorial.math.lamar.edu
6. Khan Academy. (n.d.). Precalculus. khanacademy.org
7. Math is Fun. (n.d.). Vectors. mathsisfun.com

If conic sections look like a grab-bag of formulas, let me offer a simpler way in. All four shapes, the circle, the ellipse, the parabola, and the hyperbola, come from one simple act: slicing a cone at different angles. And each shape announces itself in its equation with a small, readable clue. Once you learn to spot the clue, naming a conic takes about two seconds. We will meet the shapes one at a time.

The big picture

Four curves, one family. Tilt a slice through a cone and you get a circle, an ellipse, a parabola, or a hyperbola. Each has a standard equation that lets you read off its center, radius, or other landmarks without plotting a single point. These shapes are everywhere, from planetary orbits to satellite dishes, so reading them quickly is a real skill.

Key idea: the squared terms in the equation tell you which conic it is; the standard form then hands you the key features.

Circles

A circle is every point a fixed distance, the radius r, from a center (h, k):

(x − h)² + (y − k)² = r²

Read (x − 2)² + (y + 3)² = 25 slowly: the center is (2, −3) and the radius is √25 = 5. Watch the sign flip, since (y + 3) is really (y − (−3)), so k = −3.

Parabolas

A parabola is a single U-shaped curve, the set of points the same distance from a focus point and a directrix line. A vertical one with vertex (h, k) looks like y = a(x − h)² + k. The vertex is the turning point, and the sign of a says which way it opens. For y = 2(x − 1)² − 4: the vertex is (1, −4), and since a = 2 is positive, it opens upward.

Ellipses

An ellipse is an oval, the set of points whose distances to two foci add to a constant. Centered at the origin, its standard form is:

x²/a² + y²/b² = 1

The curve reaches a units left and right and b units up and down. For x²/25 + y²/9 = 1: it stretches to x = ±5 and y = ±3, so it is wider than it is tall.

Hyperbolas

A hyperbola has two separate branches. Centered at the origin, one standard form is x²/a² − y²/b² = 1. The minus sign between the squared terms is the single clue that separates a hyperbola from an ellipse.

Naming a conic at a glance

Clue in the equation	Conic
x squared and y squared, equal coefficients, added	Circle
only one variable is squared	Parabola
x squared and y squared, different positive coefficients, added	Ellipse
x squared and y squared subtracted	Hyperbola

These clues let you name a conic before doing any other work, then jump straight to the matching standard form for its features.

Let us complete the square together, slowly

Conics often arrive expanded and in disguise. Rebuild the standard form by completing the square in x and y separately. Take x² + y² − 6x + 4y − 12 = 0.

1. Group the x-terms and the y-terms, and move the number to the right: (x² − 6x) + (y² + 4y) = 12.
2. Complete the x-square: half of −6 is −3, and (−3)² = 9. (Half the coefficient, then squared.)
3. Complete the y-square: half of 4 is 2, and 2² = 4.
4. Add both to each side to stay balanced: (x² − 6x + 9) + (y² + 4y + 4) = 12 + 9 + 4.
5. Write each group as a square: (x − 3)² + (y + 2)² = 25.

The mystery equation was a circle all along, with center (3, −2) and radius √25 = 5. Every expanded conic yields to this same regrouping.

Try it: give the center and radius of (x + 4)² + (y − 1)² = 49, and name the conic x²/16 + y²/4 = 1.

Answer: Center (−4, 1) (flip the signs inside), radius √49 = 7. The second equation has both terms squared and added with different bottoms, so it is an ellipse. Well done.

Foci of an ellipse

The two foci sit inside the ellipse on its longer axis, at distance c from the center where c² = a² − b² (here a is the square root of the larger bottom). For x²/25 + y²/9 = 1: c² = 25 − 9 = 16, so c = 4 and the foci are at (4, 0) and (−4, 0). The defining property, distances to the two foci adding to a constant (here 2a = 10), is exactly what makes an ellipse the shape of a planet's orbit.

Worked example: from a messy equation to a standard form

Real equations arrive expanded. Completing the square on both variables reveals which conic you actually have.

Question: Identify and describe 9x² + 4y² − 36x + 8y + 4 = 0.

1. Group and move the constant. (9x² − 36x) + (4y² + 8y) = -4.
2. Factor out the leading coefficients. 9(x² − 4x) + 4(y² + 2y) = -4.
3. Complete each square. Half of -4 is -2, squared is 4. Half of 2 is 1, squared is 1. Add them inside, and add the matching amounts to the right: 9(4) = 36 and 4(1) = 4.
4. 9(x − 2)² + 4(y + 1)² = -4 + 36 + 4 = 36.
5. Divide by 36 to reach standard form. (x − 2)²/4 + (y + 1)²/9 = 1.

Read it off. Both squares are positive and added, so this is an ellipse, centred at (2, -1). The bigger denominator, 9, sits under y, so the major axis is vertical with a = 3; the minor axis is horizontal with b = 2. The ellipse reaches 3 units up and down from the centre, and 2 units left and right, so it spans from (2, -4) to (2, 2) vertically and (0, -1) to (4, -1) horizontally.

Foci. For an ellipse, c² = a² − b² = 9 − 4 = 5, so c = √5 ≈ 2.24 along the major (vertical) axis. The foci are at (2, -1 ± 2.24).

Check a point. The topmost point should be (2, 2). Substitute into the standard form: 0/4 + 9/9 = 1. It satisfies the equation, so the centre and the value of a are right.

Where people get stuck on step 3: forgetting that the number added inside the bracket gets multiplied by the factor outside before it can be added to the other side. We added 4 inside a bracket multiplied by 9, so 36 - not 4 - went to the right. Getting that wrong shifts the whole conic and is the single most common error in this procedure.

Asymptotes of a hyperbola

A hyperbola x²/a² − y²/b² = 1 opens left and right and hugs two crossing guide lines, its asymptotes y = (b/a)x and y = −(b/a)x. For x²/9 − y²/4 = 1: a = 3 and b = 2, so the asymptotes are y = ±(2/3)x and the vertices are (±3, 0). Sketch the guide lines first and the branches almost draw themselves.

The focus of a parabola

A vertical parabola with vertex at the origin can be written x² = 4py, where p is the distance from the vertex to the focus (and, on the far side, to the directrix). For x² = 8y: 4p = 8 gives p = 2, so the focus is (0, 2) and the directrix is the line y = −2. Every point on the curve stays the same distance from that point and that line.

Where people get stuck

The most frequent slip is the sign inside a circle or parabola. The standard form uses (x − h), so (x + 4) means h = −4, not +4. Read it as "x minus negative four" and the sign sorts itself out. The second common error is reading the number on the right of a circle as the radius; that number is r², so take its square root. And when you cannot decide ellipse or hyperbola, look only at the sign between the squared terms: a plus is an ellipse, a minus is a hyperbola.

Real-world reason this matters

Each conic earns its keep through a reflection trick. A parabolic dish gathers incoming parallel signals to its focus, where the receiver sits, and a headlight runs the same trick in reverse. An elliptical whispering gallery carries a murmur from one focus to the other, and the same focusing sends planets around the Sun, which sits at one focus. Hyperbolas shape long-range navigation systems, the ground curve of a sonic boom, and the graceful waist of a cooling tower.

Common misconceptions

• "(x + 4) means the center is at +4." The standard form uses (x - h), so (x + 4) means h = -4; the sign flips.
• "The radius is the number on the right side." The right side is r squared; take its square root.
• "In an ellipse, a always goes with x." The larger bottom is a squared wherever it sits; if it is under y, the ellipse is tall rather than wide.
• "Ellipses and hyperbolas are basically the same." The sign between the squared terms is the whole story: plus is an ellipse, minus is a hyperbola.
• "Completing the square moves the graph." Adding the same amount to both sides only rewrites the equation; the curve never moves.

Recap

The four conic sections come from slicing a cone and are named by their squared terms: both squared with equal coefficients (circle), both squared and added with different coefficients (ellipse), one squared (parabola), squared terms subtracted (hyperbola). Standard forms expose the features: center and radius for a circle, vertices and foci with c² = a² − b² for an ellipse, vertex and focus with x² = 4py for a parabola, and vertices with asymptotes y = ±(b/a)x for a hyperbola. Completing the square turns any expanded equation back into standard form.

Sources

1. OpenStax. (2021). 10.1 The ellipse. In Precalculus 2e. openstax.org
2. OpenStax. (2021). 10.2 The hyperbola. In Precalculus 2e. openstax.org
3. OpenStax. (2021). 10.3 The parabola. In Precalculus 2e. openstax.org
4. OpenStax. (2021). 12.1 The ellipse. In Algebra and Trigonometry 2e. openstax.org
5. Dawkins, P. (n.d.). Ellipses. Paul's Online Math Notes. Lamar University. tutorial.math.lamar.edu
6. Dawkins, P. (n.d.). Hyperbolas. Paul's Online Math Notes. Lamar University. tutorial.math.lamar.edu
7. Math is Fun. (n.d.). Conic sections. mathsisfun.com

If polar coordinates feel like a strange new language, here is a comforting thought: you already speak it. Whenever you say "the coffee shop is about two blocks that way," you are giving a distance and a direction, and that is exactly what a polar coordinate is. The usual grid says "go right this much, then up this much." Polar says "face this direction, then walk this far." Same destination, different instructions. We will translate between the two slowly.

The big picture

There are two natural ways to pin down a point. The grid way gives a right amount and an up amount. The polar way gives a distance and an angle. Neither is more correct; some problems are far easier in one than the other, especially anything with circular symmetry, like radar or a spinning antenna.

Key idea: polar coordinates locate a point by a distance r and an angle θ, and two simple formulas translate to and from the grid.

A different address system

The grid coordinates (x, y) say how far right and how far up. Polar coordinates (r, θ) instead give a distance r from the origin (here called the pole) and an angle θ measured from the positive x-axis. Reading "(4, 60°)" out loud, it says "go 4 units in the 60 degree direction."

Plotting a polar point

To plot (4, 60°): face along the positive x-axis, turn 60 degrees counterclockwise, and walk 4 units. If r is negative, walk backward instead: (−2, 30°) faces the 30 degree direction but steps 2 units the opposite way.

Let us convert one together, slowly

The bridge is basic right-triangle trig: x = r cosθ and y = r sinθ. Convert (4, 60°).

1. Find x: x = 4 cos 60° = 4 × 0.5 = 2. (Distance times the cosine.)
2. Find y: y = 4 sin 60° = 4 × (√3/2) = 2√3 ≈ 3.46. (Distance times the sine.)

So the grid point is (2, 2√3). The angle set the direction; the distance stretched it out.

Grid to polar, step by step

Going the other way uses r = √(x² + y²) and tanθ = y ÷ x. Convert (3, 3).

1. Find the distance: r = √(3² + 3²) = √18 = 3√2. (Pythagorean theorem.)
2. Find the angle: tanθ = 3 ÷ 3 = 1, and the point is in Quadrant I, so θ = 45°.

So the polar form is (3√2, 45°). Nicely done.

Try it: convert the polar point (2, 90°) to grid coordinates.

Answer: x = 2 cos 90° = 0 and y = 2 sin 90° = 2, so the point is (0, 2), straight up. Well done.

Simple polar equations

Some curves are much simpler in polar form:

• r = 5 is a circle of radius 5 around the pole, because every point sits 5 units out, whatever the angle.
• θ = 45° is a line through the origin at that fixed angle.
• r = 2 + 2 cosθ traces a heart-shaped curve called a cardioid, which would be clumsy to write in x and y.

Choosing coordinates that match a problem's symmetry, circular symmetry especially, can turn a hard equation into an easy one. That is the practical reason polar coordinates matter.

Many names for one point

Unlike grid coordinates, polar names are not unique. The same point answers to (r, θ), to (r, θ + 360°) (spin an extra full turn), and to (−r, θ + 180°) (face the opposite way and walk backward). For instance, (3, 45°), (3, 405°), and (−3, 225°) all mark the same spot. Exams love this fact, so it is worth seeing once here.

A negative-r conversion

Convert (−2, 30°) to grid coordinates. The formulas do not care about the sign of r.

1. x = r cosθ = −2 cos 30° = −2 × (√3/2) = −√3 ≈ −1.73.
2. y = r sinθ = −2 sin 30° = −2 × 0.5 = −1.

The point is (−√3, −1), sitting in Quadrant III, exactly opposite the 30 degree direction, just as a negative radius promised.

Converting equations, not just points

Convert the polar equation r = 4 cosθ to grid form. The trick is to multiply both sides by r so every piece becomes convertible.

1. Multiply by r: r² = 4 r cosθ.
2. Swap in r² = x² + y² and r cosθ = x: x² + y² = 4x.
3. Complete the square: x² − 4x + y² = 0 becomes (x − 2)² + y² = 4.

It is a circle of radius 2 centered at (2, 0). Going the easy direction, the circle x² + y² = 9 is simply r = 3.

Worked example: a polar equation becomes a familiar circle

Some polar equations look exotic and turn out to be shapes you have known for years. The tools are always the same three bridges: x = r cosθ, y = r sinθ, and r² = x² + y².

Question: What curve is r = 4 cosθ?

1. Multiply both sides by r. This is the key move, because it manufactures the r² and r cosθ that the bridges recognize: r² = 4r cosθ.
2. Translate. r² becomes x² + y², and r cosθ becomes x: x² + y² = 4x.
3. Gather and complete the square in x. x² − 4x + y² = 0, then add 4 to both sides: (x − 2)² + y² = 4.
4. Read it. A circle of radius 2 centred at (2, 0) - passing through the origin and through (4, 0).

Check with a few angles. At θ = 0: r = 4 cos 0 = 4, giving the point (4, 0). At θ = π/2: r = 4 cos(π/2) = 0, the origin. At θ = π/3: r = 4(0.5) = 2, and that point in rectangular form is (2 cos 60°, 2 sin 60°) = (1, 1.732). Test it: (1 − 2)² + 1.732² = 1 + 3 = 4. It sits on the circle.

Notice a pretty fact: the whole circle is traced as θ runs from 0 to π, not 2π. Between π and 2π, cosine goes negative, r goes negative, and the negative radius points those angles straight back onto the same circle - drawing it a second time.

Quadrant care going back to polar

Convert (−3, 3) to polar. The distance is r = √(9 + 9) = 3√2. Naively, tanθ = 3 ÷ (−3) = −1 suggests θ = −45°, but the point sits in Quadrant II (x negative, y positive), so add 180 degrees: θ = 135°. The polar form is (3√2, 135°). Always picture the point before trusting the inverse-tangent button.

Where people get stuck

The biggest trap is trusting the calculator's inverse tangent without checking the quadrant. That button only knows the ratio y ÷ x, so it cannot tell Quadrant II from Quadrant IV; whenever x is negative, add 180 degrees. The second surprise is that one point has many polar names, through extra full turns and negative radii, which is a feature rather than an error. And a quiet one: make sure the calculator is in the right mode, degrees or radians, or every conversion silently comes out wrong.

Real-world reason this matters

Anything measured by distance and bearing is already polar: radar and sonar screens, air-traffic scopes, and compass navigation. A directional microphone advertises a cardioid pickup pattern, literally the polar graph r = a + a cosθ, strong in front and nearly dead behind. Antenna engineers publish radiation patterns as polar plots, and robotic arms and CNC machines often think in radius-and-angle terms around a rotation point.

Common misconceptions

• "Each point has one polar name." Every point has infinitely many, through extra turns and negative radii.
• "r must be positive." A negative r is legal and means walk backward along the opposite ray.
• "theta is always the inverse tangent of y over x." The inverse tangent needs a quadrant adjustment whenever x is negative.
• "Polar and grid graphs cannot describe the same curve." They are two languages for one geometry; convert with the standard bridges.
• "Degree and radian mode do not matter here." Evaluating r cos theta in the wrong mode quietly ruins every conversion.

Recap

Polar coordinates locate a point by distance r from the pole and angle θ from the positive x-axis. Convert with x = r cosθ, y = r sinθ, r = √(x² + y²), and tanθ = y ÷ x plus a quadrant check. Points have many polar names, including negative-r versions. Equations convert too, often by multiplying through by r, and the classic polar families (circles, cardioids, roses, limacons) model real radar and microphone patterns. You can now move between the two systems with confidence.

Sources

1. OpenStax. (2021). 8.3 Polar coordinates. In Precalculus 2e. openstax.org
2. OpenStax. (2021). 8.4 Polar coordinates: Graphs. In Precalculus 2e. openstax.org
3. OpenStax. (2021). 8.5 Polar form of complex numbers. In Precalculus 2e. openstax.org
4. OpenStax. (2021). 10.3 Polar coordinates. In Algebra and Trigonometry 2e. openstax.org
5. Dawkins, P. (n.d.). Polar coordinates. Paul's Online Math Notes. Lamar University. tutorial.math.lamar.edu
6. Khan Academy. (n.d.). Precalculus. khanacademy.org
7. Math is Fun. (n.d.). Polar and Cartesian coordinates. mathsisfun.com

You have reached the doorway to calculus, and if that word has ever sounded out of reach, here is the good news: the idea waiting behind it is gentle and almost obvious once you see it. A limit answers a simple question, "where is this function heading?" Not where it lands exactly, but where it is clearly going as you get closer and closer. If you have ever watched a car slow toward a stop sign and known where it would end up before it got there, you already have the instinct. We will build it up slowly.

The big picture

A limit is about approach, not arrival. It asks what value a function is closing in on as the input closes in on some number, even if the function never quite reaches it, or is not defined there at all. This one idea is the foundation under every piece of calculus that follows.

Key idea: a limit is the value a function heads toward as x heads toward a point, no matter what happens exactly at that point.

The idea in words

We write lim (x approaches c) f(x) = L, read "the limit, as x approaches c, of f of x, equals L." It means: as x gets closer and closer to c, the output f(x) gets closer and closer to L. The whole focus is on the journey toward c, never on the single instant at c.

Let us sneak up on one together, slowly

Look at f(x) = (x² − 1) ÷ (x − 1) near x = 1. Putting in x = 1 directly gives 0 ÷ 0, which is undefined, so instead we creep toward 1 from both sides and watch:

x	0.9	0.99	1.01	1.1
f(x)	1.9	1.99	2.01	2.1

From both sides the outputs close in on 2. So lim (x approaches 1) f(x) = 2, even though the function has a hole exactly at x = 1. The function never gets to 2 there, yet it is unmistakably heading for 2.

Confirming with algebra

Factor the top: x² − 1 = (x − 1)(x + 1).

1. Rewrite: f(x) = (x − 1)(x + 1) ÷ (x − 1). (Factored form.)
2. For every x ≠ 1, the (x − 1) on top and bottom cancel, leaving f(x) = x + 1.
3. As x approaches 1, x + 1 approaches 2.

The algebra matches the table exactly. This is our first example of clearing an indeterminate form (the 0/0 shape) by simplifying first.

When you can just substitute

If a function has no break at c, like any polynomial, you can simply plug in. lim (x approaches 3) (2x + 1) = 2(3) + 1 = 7, because the line flows smoothly through that point with no hole and no jump.

Try it: find lim (x approaches 2) (x² − 4) ÷ (x − 2).

Answer: Factor the top: (x − 2)(x + 2) ÷ (x − 2) = x + 2. As x approaches 2, this approaches 4. Well done, that was a 0/0 form and you cleared it.

One-sided limits

Sometimes a function heads toward different values from the left and from the right. If the left-hand limit and the right-hand limit disagree, the two-sided limit does not exist. Picture a step that jumps from 2 up to 5 at x = 0: coming from the left it heads to 2, from the right it heads to 5, so lim (x approaches 0) does not exist. Both sides have to agree.

Limit laws

Limits play nicely with arithmetic: the limit of a sum is the sum of the limits, and the same holds for differences, products, constant multiples, and quotients (as long as the bottom limit is not zero). These limit laws are why substitution works for polynomials: lim (x approaches 2) (x² + 3x) = 2² + 3(2) = 4 + 6 = 10, computed one piece at a time.

Another 0/0 rescue

Evaluate lim (x approaches 2) (x² − 5x + 6) ÷ (x − 2).

1. Try substitution: (4 − 10 + 6) ÷ 0 = 0 ÷ 0, indeterminate, so simplify first.
2. Factor the top: x² − 5x + 6 = (x − 2)(x − 3).
3. Cancel the common (x − 2): for x ≠ 2 the function is x − 3.
4. Now substitute: 2 − 3 = −1.

So the limit is −1. A table agrees: at 1.99 the value is −1.01, at 2.01 it is −0.99, both closing in on −1.

Limits at infinity

We can also ask where a function heads as x grows without bound, written lim (x approaches infinity). The key fact is lim (x approaches infinity) (1 ÷ x) = 0: one divided by a huge number is nearly nothing. For a ratio of polynomials, divide top and bottom by the highest power: lim (x approaches infinity) (3x² + 1) ÷ (x² + 2) becomes (3 + 1/x²) ÷ (1 + 2/x²), the tiny fractions vanish, and it settles to 3 ÷ 1 = 3. That is exactly the horizontal-asymptote rule from the rational-functions lesson, seen from the inside.

When a limit does not exist

Not every limit has an answer, and knowing the three ways one can fail is as useful as computing the ones that work.

Failure 1: the two sides disagree. Consider f(x) = |x| / x near 0. For any positive x, |x| = x, so the fraction is 1. For any negative x, |x| = -x, so the fraction is -1.

x	-0.1	-0.01	-0.001	0.001	0.01	0.1
f(x)	-1	-1	-1	1	1	1

The left-hand limit is -1 and the right-hand limit is 1. Since a limit must be one single number that both sides agree on, this limit does not exist. The graph has a jump at 0, and no amount of zooming closes it.

Failure 2: the values blow up. For g(x) = 1/x² near 0, plug in 0.1 to get 100, then 0.01 to get 10 000, then 0.001 to get 1 000 000. The values do not settle on anything; they grow without bound. We write lim = infinity as a description of the behavior, but strictly the limit does not exist, because infinity is not a number.

Failure 3: endless oscillation. Some functions, such as sin(1/x) near 0, swing between -1 and 1 faster and faster as you approach, never settling. No zoom level shows a trend.

Where people get stuck: concluding a limit exists after checking only one side, or only a couple of values. The safe habit is to approach from both directions and watch the numbers converge. If the left column and the right column head to different places - or nowhere at all - say clearly that the limit does not exist rather than picking one of the two.

One reassurance: a limit failing to exist is not an error in your work. It is a genuine, reportable feature of the function, and on an exam "the limit does not exist because the left-hand and right-hand limits differ" is a complete and correct answer.

Continuity

A function is continuous at a point c when three things agree: f(c) is defined, the limit as x approaches c exists, and the two are equal. In plain words, you can draw the graph through that point without lifting your pencil. The hole in (x² − 1) ÷ (x − 1) is a removable discontinuity (the limit exists, the value is just missing); a step function has a jump discontinuity; and 1 ÷ x at 0 has an infinite discontinuity. Polynomials are continuous everywhere, which is precisely why substitution works for them.

A first taste of the derivative

Here is where all of this is heading. Take the parabola f(x) = x² and find the slope of the line through (1, 1) and a nearby point (1 + h, (1 + h)²).

1. Slope is rise over run: ((1 + h)² − 1) ÷ h.
2. Expand the top: (1 + 2h + h² − 1) ÷ h = (2h + h²) ÷ h.
3. Factor and cancel one h (fair, since h is tiny but not zero): 2 + h.
4. Let h approach 0: the slope approaches 2.

So the parabola has an exact steepness of 2 at the point (1, 1). We found it by taking a limit of a 0/0 form and clearing it with algebra, the very skills of this lesson. That number is called the derivative, and it is the opening act of calculus. Look how far you have come.

Where people get stuck

The most common misunderstanding is thinking the limit must equal the function's value at the point. It does not; the limit is about the approach, and the function can be undefined there or even defined with a different value. The second is stopping at 0 ÷ 0. That is not a dead end, it is a signal to factor and simplify, after which a clean finite answer often appears. And remember that a limit "equal to infinity" is really shorthand for growing without bound; as an ordinary number, that limit does not exist.

Real-world reason this matters

A speedometer is a limit in action: average speeds over shorter and shorter time intervals, approaching the instantaneous speed at one moment. Drug concentrations approach a steady level (a limit at infinity), populations approach a carrying capacity, and interest compounded more and more often approaches the continuous limit tied to the number e. Whenever a process settles, saturates, or needs an exact instantaneous rate, a limit is quietly doing the work.

Common misconceptions

• "The limit is the value of the function at the point." The limit describes the approach; the function may be undefined there, or defined with a different value.
• "0/0 means the limit does not exist." It means the form is indeterminate; algebra often reveals a perfectly finite limit.
• "If f(c) is undefined, the limit cannot exist." Holes are the classic counterexample: undefined value, existing limit.
• "A limit of infinity is an ordinary answer." It reports unbounded growth; as a finite number the limit does not exist.
• "Checking one side is enough." The two-sided limit needs the left and right approaches to agree.

Recap

A limit is the value a function approaches as its input approaches a point, independent of the value at the point. Estimate limits with tables and graphs, compute them by substitution when the function is continuous, and rescue 0/0 forms by factoring and canceling. One-sided limits must agree for a two-sided limit to exist, limits at infinity explain horizontal asymptotes, infinite limits explain vertical ones, and continuity means the limit equals the value. The derivative, a limit of slopes, is where calculus begins, and you just met it.

Sources

1. OpenStax. (2021). 12.1 Finding limits: Numerical and graphical approaches. In Precalculus 2e. openstax.org
2. OpenStax. (2021). 12.2 Finding limits: Properties of limits. In Precalculus 2e. openstax.org
3. OpenStax. (2021). 12.3 Continuity. In Precalculus 2e. openstax.org
4. OpenStax. (2021). 12.4 Derivatives. In Precalculus 2e. openstax.org
5. Dawkins, P. (n.d.). Computing limits. Paul's Online Math Notes. Lamar University. tutorial.math.lamar.edu
6. Dawkins, P. (n.d.). Continuity. Paul's Online Math Notes. Lamar University. tutorial.math.lamar.edu
7. Khan Academy. (n.d.). Precalculus. khanacademy.org