Algebra Unit A5
Linear Functions
Function notation, and what slope and intercept mean out in the world.
Naming the machine — f(x) notation, evaluating and solving, rates from tables, and interpreting slope and intercept in context.
The machine finally gets a name
The phone-bill machine has been with you since F8: $ plus $ per gigabyte. A1 ran it backwards; A4 drew every month of it at once as . Now put a second plan next to it — $ up front but $ per gigabyte — and watch the old notation buckle: ”, and also , and the first at is less than the second at… ” You can’t even ask which plan is cheaper without pointing awkwardly. The fix is to give each machine a name:
That’s all function notation is: a name for a rule, with a slot for the input. Read as ” of ”: the machine is called , you drop a number of gigabytes into the slot, and out comes a bill. means “the output of when the input is ” — a single number, computed by substituting: . Now the comparison that was unaskable is one clean line: is ? ( versus — the pricier-per-gig plan wins at four gigs.)
Two questions you can ask a machine
Everything in this unit is one of two moves. Evaluating feeds the machine an input:
Solving hands the machine a target output and asks which input produced it. Which month costs exactly $? That’s the equation , i.e. — and A1 already taught you the undo order: subtract the first (), then divide by (). Forward, the machine multiplies then adds; backwards, you subtract then divide. Same machine, two directions.
Predict before you look: with , what comes out for input ? Then flip to solve and hand it target — watch the undo boxes run in reverse order. Now set , keep , and ask for target : why does a flat machine have no answer for that? (And which single target has every answer?)
The parentheses are a slot, not a product
Here’s the notation’s one nasty trick. Everywhere else in algebra, means multiply — so your eye wants to mean . If that were true, would be for every function in the world; but — an empty month still costs the base fee. And doubling the input would double the output; but , nowhere near . The parentheses after a function name are a mail slot: is “what returns for ,” one number, and there is no floating free to multiply or cancel with anything.
Every fact about is a point
A4 called a two-variable equation a membership test for points. Function notation is the same test, written tighter: saying and saying “the graph of passes through ” are the same sentence. Input across, output up. Three special cases do most of the SAT’s work: is the y-intercept (the start), solving is finding where the line reaches height , and solving finds the x-intercept — where the graph touches the floor.
Reading a machine off a table
Often you’re handed no rule at all — just values. Say a table shows , , . Is it linear? Check the steps: each time the input climbs by , the output climbs by . Steady steps in, steady steps out — that’s the linear signature. The rate, though, is per one unit of input:
The eye-catching number is the output step, — but the inputs move by , so the slope is , not . (Tables step by twos and fives precisely to set this trap.) With the rate and the start, the rule reassembles itself: , and now is a substitution instead of a fifty-row table.
What the numbers mean out in the world
The SAT’s favorite function question contains no algebra at all: “In , what does the represent?” The answer is always the same two sentences, and units decide everything. The slope is a rate, so its meaning needs the word per: $ per gigabyte. The intercept is a value at zero, the reading before anything happens: $ when , the base fee. Swap them and the sentence turns nonsense — a bill can’t start at "" or grow by "" flat. When the slope is negative — a candle — the rate reads as a loss (“burns cm each hour”), and one more landmark gains a meaning: the x-intercept is the runs-out moment, at .
Pick the candle. Before touching anything, predict: raising the burn rate moves which end of the line — and does the “gone at…” marker slide left or right? Then slide only the starting length: the tilt never changes, the whole story just starts higher (A4’s -slider, now with a plot). The scooter’s unlock fee does the same thing in dollars.
The one thing to remember
A function is a named machine: , so every fact is a point on its graph. Evaluating substitutes forward; solving runs the machine backwards. And in any real story, the slope is the rate (the “per” number), the y-intercept is the starting value, and the x-intercept is when it runs out.
Reading the notation
| You see | It says | You do |
|---|---|---|
| a rule named | nothing — it’s a definition | |
| the output at input | substitute: | |
| “which input gives ?“ | solve | |
| the value at zero | the y-intercept, | |
| point is on the graph | input across, output up |
From a table of values
- Linear check: equal input steps ⇒ equal output steps.
- Rate: — divide by the input step, which is often or , not .
- Start: (or work backwards to it). Then .
Interpreting a model
| Number | Meaning in context | Units |
|---|---|---|
| slope | rate of change — the “per” number | y-units per x-unit |
| y-intercept | starting value, when | y-units |
| x-intercept | the “runs out” moment () | x-units |