Turns the loose word 'function' into a test you can run: every input has exactly one output. Builds function notation as a name for an output rather than a multiplication, separates evaluating f(3) from solving f(x) = 3, and introduces the vertical and horizontal line tests as the same definition seen from two directions.
Subject: Precalculus · 71 slides · symbolic lesson
Open the interactive version of this deck
Title
Precalculus · Section 1.1
Turning a loose word into a test you can actually run
Objectives
You have used the word function since Algebra 1, usually to mean a formula with an x in it. This section replaces that habit with a test, and gives you the notation the whole course is written in.
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, pp. 9-33 — learning objectives 1.1.1 to 1.1.5
Almost everything in this course is going to be said in the language of functions, so this first section is worth more care than its easy-looking content suggests. You probably already have a working idea of what a function is, something like a formula with an x in it, or a rule that does something to a number. That idea is not wrong, but it is too vague to decide the hard cases, and the hard cases are what the exercises ask about.
So the section replaces the vague idea with a test you can run on any object at all: a list of pairs, a table, a graph, a menu, a grading scheme. The test is one sentence long, and the whole first part of the lesson is about applying it honestly, including in the direction you were not expecting.
The rest is notation and habits. You will learn to read and write the shorthand that names a rule and its output, to tell apart the two tasks that look identical on the page, and to run two sweeps across a graph that answer two different questions. None of it is hard; most of it is easy to do carelessly, which is why the lesson keeps stopping to show you how the careless version fails.
Warm-up
Discussion prompt
One machine takes a person's name and returns their date of birth. A second machine takes a date of birth and returns a person's name. Only one of these can be relied on to give a single answer. Which one, and what exactly goes wrong with the other?
Commit to an answer in writing before you reveal anything. This one question contains the whole definition, and it is much more memorable when you have discovered it yourself than when you have read it.
The point is not that birthdays are special. It is that the reliability of a rule depends on which way round you run it. Plenty of pairings are perfectly dependable in one direction and useless in the other, and the same pairing can therefore be a function or not a function depending on which quantity you decided was the input.
Keep that idea with you for the next ten slides. Every example in the first part of this lesson is a pairing looked at from both directions, and in most of them exactly one of the two directions works.
Section
Part 1
Concept
Start with something weaker than a function. A relation is any collection of input-output pairs at all, with no rules about it.
\[ \{(1,2),\;(2,4),\;(3,6),\;(4,8),\;(5,10)\} \]
domain — The set of first components: every value that is allowed as an input.
range — The set of second components: every value that actually comes out as an output.
\[ \text{domain} = \{1,2,3,4,5\}, \qquad \text{range} = \{2,4,6,8,10\} \]
The input is also called the independent variable and is usually written x. The output is the dependent variable and is usually written y.
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, p. 9 — relations, domain and range
Start with the weaker word. A relation is any set of pairs whatsoever, with no conditions attached, so every function is a relation but most relations are not functions. Having the weaker word available is what lets the definition of a function be stated as a restriction rather than as a description.
The two sets that come with a relation have names you will use constantly. The domain is everything allowed in as an input, and the range is everything that actually comes out. Notice that the range is what comes out, not what could come out: in the example the range is the five even numbers up to ten, not all the even numbers.
The habit of writing the input first in every pair is a convention, but it is a load-bearing one. If you write a pair the other way round by accident, every test in this lesson will give you the wrong answer, because every test looks at the first coordinate.
Concept
function — A relation in which each possible input value leads to exactly one output value. The output is a function of the input.
The word exactly is doing all the work. Not at most one, which would let an input have no answer. Not at least one, which would let an input have two.
\[ \text{for each } x \text{ in the domain, exactly one } y \]
Now take the pairs that attach the words even and odd to the first five whole numbers. Written as pairs with the word first, the input odd appears three times.
\[ \{(\text{odd},1),\;(\text{even},2),\;(\text{odd},3),\;(\text{even},4),\;(\text{odd},5)\} \]
\[ \text{odd} \longmapsto 1, \; 3, \; 5 \quad \Longrightarrow \quad \text{not a function} \]
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, pp. 9-10 — the definition of a function
Read the definition slowly and notice where the pressure sits. It is on the word exactly. If the rule only promised at least one output, you could not write down a symbol for the output and expect it to name one number, and all the algebra you are about to do would be meaningless.
The even and odd example is worth working through yourself rather than reading. When you write the pairs with the word first, the input odd shows up three separate times, attached to 1, then 3, then 5. That is one input with three different outputs, so the definition rejects it immediately.
Notice what would happen if you wrote those pairs the other way round, with the number first. Then each of the five numbers appears once, and the relation would be a function. Same data, opposite verdict, because the direction changed. Always settle the direction before you run the test.
Picture it
Figure (svg): Three mapping diagrams. In (a) three inputs p, q and r send arrows to two outputs m and n, with q and r both landing on n. In (b) three inputs each send one arrow to their own output. In (c) the input q sends two arrows, to y and to z.
Cover every label and ask one question of each diagram: does any single input have two arrows leaving it? Diagram (a) has two arrows arriving at n, which is allowed. Diagram (c) has two arrows leaving q, which is not.
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, p. 10 — Figure 1.2
Take a moment with these three pictures, because they are the definition in a form you can check at a glance. Each oval on the left holds the inputs and each oval on the right holds the outputs, and every arrow is one pairing.
Cover the captions and ask one question of each diagram: does any single dot on the left have two arrows leaving it? In diagram (a) the answer is no, even though two arrows arrive at the same place on the right. In diagram (b) the answer is also no, and here nothing is crowded on either side. In diagram (c) the input q has two arrows leaving it, and that is the only thing on this slide that breaks the rule.
Diagram (b) is quietly the most important one. It satisfies a stronger condition than the other two, and at the end of the lesson that condition gets its own name and its own test. For now, just notice that it looks different.
Notation
Annotate
On: \( f: x \longmapsto f(x), \qquad \text{each } x \text{ has exactly one } f(x) \)
Work through the annotations one at a time and keep asking what each piece would cost you if it were missing. The name f exists so that you can talk about the rule without writing the rule out, which matters once the rule is three lines long.
The third annotation is the one that decides every exercise in the first part of this section. One arrow leaves each input. Two arrows arriving somewhere is a fact about the outputs, and the definition never mentions the outputs.
The last annotation explains why the word exactly cannot be softened. Because each input has one output and no more, the symbol f of 3 refers to one definite number, so you may add it, square it, or set it equal to something. If an input were allowed two outputs, the symbol would be ambiguous and every line of algebra after it would be too.
Intuition
The definition constrains what leaves each input and says nothing about what arrives at each output. That asymmetry is deliberate, and it is where most early marks are lost.
\[ \text{two arrows out of one input: forbidden} \]
\[ \text{two arrows into one output: perfectly legal} \]
Squaring is the standard example. Both minus 3 and 3 are sent to 9, so the output 9 is crowded, and squaring is still a function because no single input was sent to two places.
\[ (-3)^2 = 9 \quad \text{and} \quad 3^2 = 9 \]
Later in the lesson you will meet the extra condition that does forbid a crowded output. It has its own name, one-to-one, precisely because it is not part of being a function.
This slide exists because the asymmetry of the definition is the single most common source of lost marks in this section. Half of a class will reject a perfectly good function because they spotted a repeated output.
Say it to yourself in the negative form, which is easier to apply: the only thing that disqualifies a relation is one input with two different outputs. Anything else you notice is irrelevant. Repeated outputs, missing outputs, strange-looking numbers, outputs that grow quickly, none of it matters.
The squaring example is the one to keep. Minus three and three both land on nine, so the output nine is reached twice, and squaring is still a function because neither minus three nor three was sent to two places. Later in the lesson you will meet the extra condition that does forbid a crowded output. The fact that it needs a separate name, one-to-one, is the clearest evidence that it is not part of being a function.
Worked example
A coffee shop menu lists a plain donut at 1.49, a jelly donut at 1.99 and a chocolate donut at 1.99. Two questions, not one.
Take the item as the input
Why: The outputs are then the prices.
\[ \text{plain} \mapsto 1.49, \quad \text{jelly} \mapsto 1.99, \quad \text{chocolate} \mapsto 1.99 \]
Check each input for a second output
Why: Every item on the menu is listed with one price.
\[ \text{price is a function of the item} \]
Figure (svg): Two mapping panels for a coffee shop menu. On the left, each of three donuts sends one arrow to a price, with jelly and chocolate both reaching 1.99. On the right, the price 1.99 sends two arrows, one to jelly donut and one to chocolate donut.
Now swap the roles
Why: Take the price as the input and the item as the output.
\[ 1.99 \mapsto \text{jelly}, \quad 1.99 \mapsto \text{chocolate} \]
Find the repeated input
Why: The input 1.99 leads to two different outputs.
\[ \text{the item is NOT a function of the price} \]
Check the answer against the definition
Why: Only the second direction has an input with two outputs, and that is the only thing the definition forbids, so exactly one of the two directions is a function.
\[ \text{item} \to \text{price: yes}, \qquad \text{price} \to \text{item: no} \]
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, pp. 10-11 — Example 1.1
Read the question twice before starting. The example asks two questions, not one, and they have different answers, which is the whole reason it is the first example in the book.
In the first direction the item is the input. You are not asking whether the prices repeat; you are asking whether any single item is listed at two prices. No menu does that, so the first direction is a function. In the second direction the price is the input, and now the repetition matters, because the price 1.99 has to point at both the jelly donut and the chocolate donut.
Look at the picture and notice that the arrows are the same arrows in both panels, just read from the other end. Nothing about the menu changed between the two questions. What changed was which side you called the input, and that alone flipped the verdict.
Worked example
A class converts percent grades to grade points with this rule.
| percent grade | grade point average |
|---|---|
| 0-56 | 0.0 |
| 57-61 | 1.0 |
| 62-66 | 1.5 |
| 67-71 | 2.0 |
| 72-77 | 2.5 |
| 78-86 | 3.0 |
| 87-91 | 3.5 |
| 92-100 | 4.0 |
Take the percent grade as the input
Why: Every percent from 0 to 100 sits in exactly one band, and each band names one grade point average.
\[ \text{percent} = 81 \;\longmapsto\; \text{GPA} = 3.0 \]
Conclude for that direction
Why: One input, one output, every time.
\[ \text{GPA is a function of the percent grade} \]
Figure (svg): A step graph of grade point average against percent grade: eight flat horizontal segments at heights 0, 1, 1.5, 2, 2.5, 3, 3.5 and 4, with the segment from 78 to 86 at height 3 drawn in yellow.
Swap the roles
Why: Now the grade point average is the input. The band from 78 to 86 all maps to 3.0.
\[ 3.0 \;\longmapsto\; 78, 79, 80, \ldots, 86 \]
Check by counting the outputs
Why: The single input 3.0 would need nine outputs, so the reverse direction fails; nine is more than one, which is all the definition needs to reject it.
\[ \text{percent grade is NOT a function of the GPA} \]
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, pp. 11-12 — Example 1.2 and Table 1.1
This is the same two-direction question with numbers instead of donuts, and it is worth doing carefully because the graph makes the asymmetry visible in a way the menu could not.
In the first direction, pick any percent grade you like and find its band. There is exactly one band containing it, so there is exactly one grade point average, and that is all a function needs. The fact that eighty-one and eighty-five give the same answer is not a problem.
Now look at the flat yellow segment on the graph. It sits at height three and runs from seventy-eight all the way to eighty-six. Reading the graph from the vertical axis instead of the horizontal one, the single value three points at nine different percent grades. That is nine outputs for one input, so the reverse direction is not a function. The graph is a step graph precisely because whole stretches of input share an output, and every flat stretch is a place where the reverse direction fails.
Prediction
| player | rank |
|---|---|
| Babe Ruth | 1 |
| Willie Mays | 2 |
| Ty Cobb | 3 |
| Walter Johnson | 4 |
| Hank Aaron | 5 |
Predict first
Which of the two directions between player name and rank is a function?
Correct: Both directions are functions
Why: Each player has exactly one rank, so rank is a function of the name. And each rank belongs to exactly one player, so the name is a function of the rank. Nothing repeats on either side, which is rarer than it looks: the menu and the grade table each failed in one direction.
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, p. 12 — Try It 1.1 and Table 1.2
Choose an option before revealing, and do it by running the test in both directions rather than by intuition.
Going from name to rank, no player appears twice, so every name has exactly one rank. Going from rank to name, no rank appears twice either, so every rank has exactly one player. Both directions pass, which makes this table unusual in the company of the last two examples.
It is worth noticing why this one is different. In the menu two items shared a price, and in the grade table a whole band of percents shared an average. Here nothing at all is shared, and a table where nothing is shared works both ways round. That property is exactly what the last part of this lesson will call one-to-one.
Sorting
Sort into buckets
For each set of pairs, decide whether it defines y as a function of x.
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, pp. 34-37 — Exercises 6, 7, 58, 59, 60
Sort all six before you check any of them, and force yourself to look only at the first coordinate in each pair. The temptation to scan for any repeated symbol is strong and it is wrong.
Two of these are not functions, and in both cases the same input carries two different partners. In the first one the letter a is paired with b and later with c. In the fourth, the number seven is paired with eleven and later with nine.
The two deliberately tricky ones are the fifth and the sixth. In the fifth, the output seven appears three times, which is allowed and always will be. In the sixth, the pair with input four is written out twice, but both copies say the same thing, so nothing is in conflict; it is one pair listed twice, not an input with two outputs.
Trap
A set of pairs comes up on a test and the output 4 appears twice:
\[ \{(1,2),\;(3,4),\;(5,4),\;(7,8)\} \]
\[ \text{4 repeats} \Rightarrow \text{not a function?} \]
Wrong. Nothing in the definition mentions the outputs.
Scan the first coordinates only. They are 1, 3, 5 and 7, all different, so every input has exactly one output and this is a function.
\[ \text{inputs } 1,3,5,7 \text{ all distinct} \]
Adding the pair with input 3 and output 9 would break it. Adding a pair with input 9 and output 4 would not.
This is the mistake to expect from yourself under time pressure. Your eye scans the set, notices that a number appears twice, and your hand writes not a function before your brain has checked which column that number was in.
The defence is mechanical. Before deciding anything, write the first coordinates out in a row on their own and look only at that row. If they are all different, you are finished and the answer is yes, whatever the outputs are doing.
The two follow-up questions on the right are worth answering yourself. Adding an input of three with output nine breaks it, because three already has output four. Adding an input of nine with output four does not break anything, because nine is new and the crowding at output four is nobody's business.
Two truths and a lie
Eliminate the wrong options
Two of these statements about the definition are true. Rule those two out and keep the false one.
Survives elimination: C
Why: The definition is a single one-directional condition on inputs. Claims A and B both restate what it allows; claim C invents a requirement about outputs that no part of the definition mentions.
Eliminate the two you are sure of first, rather than hunting for the false one directly. The two true claims should both feel familiar from the last three slides, and if either of them still feels doubtful, that is the thing to go back over.
The false claim is the one that invents a requirement about the range. Nothing in the definition says an output has to be used at all, let alone used twice. A function can hit an output once, twice, or a hundred times, and it can miss values entirely, and none of that has any bearing on whether it is a function.
The general lesson here is about how to test a claim about a definition: try to build a counterexample. A function with three inputs and three distinct outputs immediately refutes the false claim, and it took ten seconds to think of.
Section
Part 2
Concept
To say that height depends on age, pick a letter for each quantity and a letter for the rule that connects them.
\[ h = f(a) \]
Read that as: h is f of a. The letter f names the rule, a is the input, and the whole symbol f of a is the output. It is one number with a two-part name.
Any letter can name the rule. Writing h of a instead says the same thing, with h now doing double duty as the rule's name.
\[ y = f(x), \qquad P = W(d), \qquad D = f(m) \]
Example 1.3 names the days-in-a-month rule f, so that the number of days in March is written as one short symbol.
\[ f(\text{March}) = 31 \]
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, pp. 12-13 — function notation and Example 1.3
This is the moment the course starts speaking in symbols, so slow down and read the line out loud. The sentence h equals f of a is read as h is f of a, and it says three things at once: there is a rule, it is called f, and feeding it a produces h.
One point causes confusion for weeks if it is not settled now. The letter f is the name of the rule and is not a number. You never multiply by f, and f on its own has no value. What has a value is f of a, and that value is a number like any other.
You are free to choose the letters. Using h for the rule and h for the height at the same time is legal and common, and physics does it all the time. In this book the rule and its output are given different letters so that you can always tell which one is being talked about.
Notation
Annotate
On: \( y = f(x), \qquad f(a+b) \ne f\cdot a + f\cdot b \)
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, pp. 12-13 — the brackets indicate the input, not multiplication
The first three annotations are about reading, and the last one is about a habit that will cost you marks if it survives this lesson.
Everything hinges on realising that the brackets here are punctuation. In the expression three times a bracket, the brackets mean multiply. In f of a bracket, the identical-looking brackets mean put this in the slot. There is nothing in the typography to tell them apart, which is why you have to know from context what the symbol before the bracket is: a number multiplies, a function name does not.
The fourth annotation spells out the consequence. To work out f of a plus b you must first work out a plus b, getting a single number, and then apply the rule to that one number. Doing the rule to a and to b separately and adding afterwards is a different calculation, and it almost always gives a different answer.
Picture it
Figure (svg): A flow diagram: two arrows labelled 2 and a plus h enter a rounded box labelled f, and two arrows leave it labelled 6 and the expression for f of a plus h.
Reading the notation as a machine makes substitution obvious instead of memorised. The slot accepts a number, a letter, or a whole expression, and whatever you drop in replaces every copy of the input letter in the formula.
The book raises a fair question at this point: why not write y of x, using the same letter for the rule and the output? You may, and physics often does. In a maths course it helps to keep the rule and its output as separate things, because the rule is a procedure and the output is a number.
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, p. 14 — the Q and A on writing y as a function of x
Imagining the rule as a machine is not just a friendly picture; it makes the substitution rule obvious rather than memorised. The machine has one slot, and it does not care what shape the thing you drop in is.
Trace the second arrow in particular. The expression a plus h goes into the slot as one object, so wherever the formula had the input letter, the whole expression arrives, brackets and all. Those brackets are not decoration; they are what keeps the expression together while the rule squares it or multiplies it.
The book pauses here to raise a fair question: why not use the same letter for the rule and for the output, and write y of x? You may, and applied subjects often do. In a maths course it helps to keep them apart, because the rule is a procedure you can talk about before you feed it anything, and the output is just a number.
Socratic
\[ N = f(y): \quad N \text{ is the number of police officers in year } y \]
Discussion prompt
Write out in one plain sentence what the statement f of 2005 equals 300 tells you. Then write the weight of a pig, in pounds, as a function of its age in days, using function notation.
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, p. 13 — Example 1.4 and Try It 1.2
Write both answers out in full sentences before revealing, including the units. Translating notation into English and back again is a skill that gets tested constantly in word problems, and it is usually the step people skip.
For the first part, identify the two roles before you say anything else. The thing inside the brackets is the input, so it is a year, and the thing on the right of the equals sign is the output, so it is a count of officers. Once the roles are fixed the sentence writes itself.
For the pig, the answer is not a formula and does not need to be. You are only asked to set up the notation, so you name the input, name the output, name the rule, and say what the units are. Most of the marks in questions like this are for saying clearly what each letter means.
Trap
The brackets look like the ones in ordinary algebra, so they get expanded:
\[ f(a+b) = f\cdot a + f\cdot b \]
Wrong. There is no quantity called f to distribute.
The brackets mark the slot. First work out what goes in the slot, then apply the rule to that one value.
\[ f(x) = x^2 \]
\[ f(2+3) = f(5) = 25 \]
Distributing would have given 4 plus 9, which is 13. The two answers are not equal, and the difference is not a rounding error.
\[ f(2) + f(3) = 4 + 9 = 13 \]
This is not a careless error so much as a well-trained reflex misfiring. You have spent two years expanding brackets on sight, and the notation deliberately reuses the same symbol.
The numbers on the right make the size of the error obvious. Squaring five gives twenty-five, while squaring two and three separately and adding gives thirteen. These are not nearly equal, so if you ever make this mistake in a longer problem the final answer will be badly wrong rather than slightly wrong.
There is a useful way to catch yourself. Before you expand any bracket, look at what is immediately in front of it. If it is a number or a variable, expanding is right. If it is the name of a function, expanding is meaningless, because there is no quantity there to distribute.
Matching
Match the pairs
Why: The first pair is a completed evaluation and the second is an equation waiting to be solved. The last two differ only in where h sits, and that decides whether it is added before or after the rule runs, which usually gives different numbers.
Make all four matches before checking, and pay most attention to the last two, which differ by the position of a single letter.
The first two are the two tasks you will spend the second half of this lesson separating. One of them is a completed statement of fact, with the input given and the output known. The other is a question with the output given and the input wanted, and its answer may well be a list of years.
The last two look almost identical and mean quite different things. In one, h joins the input before the rule runs, so the rule acts on the combined value. In the other, the rule runs on a by itself and h is added to the answer afterwards. Try it with a squaring rule and some small numbers; the two answers come apart immediately.
Section
Part 3
Concept
A table of values is the same information as a set of ordered pairs, laid out in two rows. The top row holds the inputs and the bottom row the outputs, so one column is one pair.
\[ \begin{array}{c|cccccccccccc} m & 1 & 2 & 3 & 4 & 5 & 6 & 7 & 8 & 9 & 10 & 11 & 12 \\ \hline D & 31 & 28 & 31 & 30 & 31 & 30 & 31 & 31 & 30 & 31 & 30 & 31 \end{array} \]
This is the days-in-a-month rule again, with the months numbered. Twelve inputs, each appearing once, so it is a function. The output 31 appears seven times, which changes nothing.
\[ D = f(m), \qquad f(2) = 28 \]
The test on a table is a scan of the top row only: does any input appear twice with different values underneath it?
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, pp. 14-15 — Tables 1.3, 1.4 and 1.5
Nothing new is happening here, which is the point. A table is the same information as a set of ordered pairs, laid out so that one column is one pair, and the test you already know applies unchanged.
Read the days-in-a-month table with the definition in mind. Twelve inputs, each appearing once, so it passes. The output thirty-one appears seven times, and by now that should feel completely unremarkable to you.
One practical point about tables. Sometimes a table is the entire function, as with the twelve months, and sometimes it is a handful of samples from something bigger, like heights measured at a few ages. The test works the same way either way, but you should know which kind you are looking at before you say anything about values that are not in the table.
Worked example
Three small tables. Run the same scan on each.
| first table | second table | third table | |
|---|---|---|---|
| inputs | 2, 5, 8 | −3, 0, 4 | 1, 5, 5 |
| outputs | 1, 3, 6 | 5, 1, 5 | 0, 2, 4 |
Scan the first table's inputs
Why: The values 2, 5 and 8 are all different.
\[ \text{first table: a function} \]
Scan the second table's inputs
Why: Minus 3, 0 and 4 are all different. The output 5 appears twice, which the definition allows.
\[ \text{second table: a function} \]
Scan the third table's inputs
Why: The input 5 appears twice, carrying 2 the first time and 4 the second.
\[ \text{third table: not a function} \]
Write the two working tables in function notation
Why: Each column becomes one statement.
\[ f(2)=1, \; f(5)=3, \; f(8)=6; \qquad g(-3)=5, \; g(0)=1, \; g(4)=5 \]
Check why the third table cannot be written that way
Why: You would have to write f of 5 equals 2 and f of 5 equals 4 at the same time, which would make one symbol name two different numbers.
\[ f(5) = 2 \text{ and } f(5) = 4: \text{ impossible} \]
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, pp. 15-16 — Example 1.5 and Tables 1.6 to 1.8
Three quick tables, and the method is the same scan each time: write out the inputs and look for a repeat that carries two different values.
The second table is the one designed to catch you. Its output five appears twice, at inputs minus three and four. That is two inputs sharing an output, which you now know is harmless, so the table is a function.
The last step of the solution is quietly important. Once a table is known to be a function you can rewrite every column in function notation, one short equation per column. The third table cannot be rewritten that way at all, and seeing why is a good test of whether the notation has landed: you would have to write the symbol f of five and then give it two different values, which would make the symbol meaningless.
Prediction
\[ \begin{array}{c|ccc} \text{input} & 1 & 2 & 3 \\ \hline \text{output} & 10 & 100 & 1000 \end{array} \]
Predict first
Does this table represent a function?
Correct: Yes: no input is repeated
Why: The inputs 1, 2 and 3 are distinct, so each has exactly one output and the table is a function. How the outputs behave is irrelevant: growth rate, size and pattern play no part in the definition. A table of three pairs is a perfectly good function with a domain of three values.
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, p. 16 — Try It 1.3 and Table 1.9
Commit to an answer first. Three of the four options are wrong for the same underlying reason, so if you picked one of them it is worth naming that reason.
The inputs are one, two and three. They are all different, so the table is a function, and that is the end of the analysis. The pattern in the outputs, the speed at which they grow, the fact that they are round numbers, none of that appears anywhere in the definition.
The option about three pairs not being enough is worth a second of thought. A function does not have to have an infinite domain. A rule defined at exactly three inputs is a perfectly respectable function whose graph is three dots.
Worked example
One table, two different tasks. Read the question carefully before you start looking.
\[ \begin{array}{c|ccccc} n & 1 & 2 & 3 & 4 & 5 \\ \hline g(n) & 8 & 6 & 7 & 6 & 8 \end{array} \]
Evaluate g at 3
Why: Find 3 in the input row and read straight down.
\[ g(3) = 7 \]
Figure (svg): A stem plot of the five values of g: g(1) is 8, g(2) is 6, g(3) is 7, g(4) is 6 and g(5) is 8, with a dashed horizontal line at height 6 touching the dots at n equals 2 and n equals 4.
Now solve the equation g of n equals 6
Why: This time 6 is the output, so search the bottom row and read up, noting every hit.
\[ g(n) = 6 \quad \text{at} \quad n = 2 \; \text{and} \; n = 4 \]
Write the answer as a list
Why: Two inputs produce the output 6, so the solution is a pair of values, not one.
\[ n = 2 \quad \text{or} \quad n = 4 \]
Check both answers in the table
Why: Reading down from 2 gives 6 and reading down from 4 gives 6, so both are genuine solutions, and nothing else in the bottom row equals 6.
\[ g(2) = 6 \;\checkmark\qquad g(4) = 6 \;\checkmark \]
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, pp. 21-22 — Example 1.11 and Tables 1.11, 1.12
This example puts the two tasks side by side deliberately, because the difference between them is the thing most often missed on a test. Read the instruction word before you touch the table.
Evaluating means the input is given, so you find three in the top row and read straight down. There is only one thing underneath it, and there could never be more than one, because that is what being a function means. One answer, guaranteed.
Solving means the output is given, so you search the bottom row instead and read upwards. Now there is no guarantee at all about how many hits you find, and here there are two. Look at the picture: the dashed line at height six touches two dots. Stopping at the first one, which is the usual error, would lose half the answer.
Discrimination
Sort into buckets
Each table has three columns. Sort them by whether the relation is a function of x.
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, pp. 37-38 — Exercises 61 to 66
Sort all six before checking, and for each one write down the repeated input, if there is one, before you decide.
Two of them fail, and in both the repeat carries two different values underneath it. Everything else passes, including the one where all three outputs are the same number and the one where the input minus two appears twice.
That last case is the one to be sure about. When an input repeats but carries the same output both times, nothing is in conflict. The relation is telling you the same fact twice, and a set does not care how many times you list the same element.
Section
Part 4
Concept
When a function is given by a formula, evaluating it is two steps: replace the input letter everywhere it appears, then simplify.
\[ f(x) = 5 - 3x^2 \]
Say it as instructions and the order becomes visible: square the input, multiply by 3, subtract the result from 5.
Write the rule once with an empty slot in place of the letter. Then whatever you drop in carries its own brackets automatically, which is where almost all the lost marks in this section live.
\[ f(\,\square\,) = 5 - 3(\square)^2 \]
\[ f(-2) = 5 - 3(-2)^2 = 5 - 3(4) = -7 \]
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, p. 16 — evaluating functions given in algebraic form
Evaluating a formula is genuinely easy, and that is exactly why it is worth building a careful habit now. The errors that appear here are not errors of understanding; they are dropped brackets and lost signs, and they happen faster when you are confident.
The habit to build is the empty slot. Copy the rule out with a box wherever the input letter appeared, before you write the number anywhere. Then drop the input into every box. Because the box already has its own brackets in your mind, the brackets arrive automatically, which is the whole trick.
Watch what those brackets do in the worked line. The input is negative, and the square has to apply to the sign as well as to the digit. Writing the input without brackets would square only the digit and leave the minus sign outside, changing the answer by a large amount. This single slip accounts for more lost marks in this section than everything else combined.
Step zero
\[ f(x) = x^2 + 3x - 4, \qquad \text{find } f(a+h) \]
Discussion prompt
Do not substitute yet. Write down how many copies of the letter x you can see in the formula, and say what each one has to become. Then predict how many pairs of brackets your first line will contain.
Write your answer down before revealing. Thirty seconds of counting here saves you from the two most common versions of this error.
Counting the copies of the input letter matters because the substitution must reach all of them. When the rule has the input in two places and you only replace one, the line still looks plausible and the answer is wrong.
Predicting the number of brackets matters for the same reason. The input here is a whole expression, so each copy needs a pair of brackets around it to hold it together while the square or the multiplication acts. If your first line has fewer pairs of brackets than the rule had copies of the letter, stop and look again.
Worked example
The same rule, fed three different things. Notice that only the first one produces a number.
\[ f(x) = x^2 + 3x - 4 \]
Substitute the number 2
Why: Every x becomes 2, in brackets.
\[ f(2) = (2)^2 + 3(2) - 4 \]
Simplify the arithmetic
Why: Square, then multiply, then combine.
\[ f(2) = 4 + 6 - 4 = 6 \]
Substitute the letter a
Why: Nothing can be simplified, because a is not a number yet. The answer is an expression, and that is a complete answer.
\[ f(a) = a^2 + 3a - 4 \]
Substitute the expression a plus h
Why: Both copies of x become the whole bracket.
\[ f(a+h) = (a+h)^2 + 3(a+h) - 4 \]
Expand the square and distribute the 3
Why: Use the square of a sum, then multiply out the second bracket.
\[ f(a+h) = a^2 + 2ah + h^2 + 3a + 3h - 4 \]
Check the expansion by putting a equal to 2 and h equal to 0
Why: The expression must collapse back to f of 2, which was 6, and it does: 4 plus 0 plus 0 plus 6 plus 0 minus 4 is 6.
\[ a=2,\; h=0: \quad 4 + 0 + 0 + 6 + 0 - 4 = 6 = f(2) \]
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, pp. 16-17 — Example 1.6 parts a, b and c
Three inputs of increasing difficulty into one rule, and the method never changes. That is the message of the example: substitution does not care what kind of thing you hand it.
The second part often feels unfinished. You substitute the letter a and then find there is nothing to simplify, so it looks as though you have not done anything. You have: the answer to the question is an expression, and an expression is a complete answer. Writing it out is the whole task.
The third part is where the brackets earn their place. Both copies of the input letter become the whole two-term expression, so the square is the square of a sum and gives three terms, not two. Expanding the square of a sum wrongly, by squaring each piece, is the classic error, and the check at the end catches it: setting h to zero must give you back the first answer, and it does.
Worked example
Now combine the last two answers. This expression is the whole reason Example 1.6 was set up that way; it becomes the derivative in Calculus I.
\[ \frac{f(a+h) - f(a)}{h}, \qquad h \ne 0 \]
Write both pieces down
Why: The two results from the previous slide.
\[ f(a+h) = a^2 + 2ah + h^2 + 3a + 3h - 4, \quad f(a) = a^2 + 3a - 4 \]
Subtract, keeping the second bracket intact
Why: The minus sign applies to all three terms of f of a.
\[ \frac{\left(a^2 + 2ah + h^2 + 3a + 3h - 4\right) - \left(a^2 + 3a - 4\right)}{h} \]
Cancel the matching terms
Why: The a squared, the 3a and the minus 4 all disappear.
\[ \frac{2ah + h^2 + 3h}{h} \]
Factor h out of the numerator
Why: Every surviving term has an h in it, which is exactly why the h in the denominator can go.
\[ \frac{h(2a + h + 3)}{h} = 2a + h + 3 \]
Check with numbers, taking a equal to 2 and h equal to 0.1
Why: The formula gives 7.1. Computing the quotient directly from f of 2.1 minus f of 2, all over 0.1, also gives 7.1.
\[ \frac{f(2.1) - f(2)}{0.1} = \frac{7.11 - 6}{0.1} = 7.1 = 2(2) + 0.1 + 3 \]
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, p. 17 — Example 1.6 part d
This expression will follow you into calculus, where it becomes the definition of the derivative, so it is worth doing slowly once rather than quickly five times.
Two things go wrong here and both are about the subtraction. The first is forgetting that the minus sign applies to all three terms of the second expression, not just to the first. Write that whole expression in brackets before you subtract anything, and then take the bracket apart deliberately.
The second is the cancellation at the end. You can only cancel the h in the denominator once every term in the numerator has an h in it, which is why the factoring step comes first. Cancelling an h out of a sum where one term does not contain one is a genuinely serious error, and doing the factoring explicitly is what stops it. The numerical check at the end is worth copying too: pick a value for a and a small value for h, compute the quotient directly, and see whether your formula agrees.
Fill the middle
\[ f(x) = x^2 + 3x - 4 \]
Fill in the blanks
f(a+h) = (a+h)^2 + 3(a+h) - 4 = a^2 + 2ah + h^2 + 3a + 3h - 4
Why: Squaring the bracket gives a squared plus twice a times h plus h squared, so the cross term is 2ah. Distributing the 3 over the second bracket gives 3a and 3h. Forgetting the cross term is the single most common slip here.
Fill both blanks before checking. Each one comes from a different step, and it is useful to know which of the two you find harder.
The first blank is the cross term from squaring a sum. Squaring a two-term bracket gives three terms, not two, and the middle one is twice the product. Missing that term is the most common error in this whole section.
The second blank comes from distributing the three across the bracket, which produces two terms, one for each part. If you wrote only one of them, look back at the original line and count how many things are inside that bracket.
Worked example
Two short ones, to make the substitution reflex automatic. The input letter changes, the method does not.
\[ h(p) = p^2 + 2p, \qquad g(m) = \sqrt{m - 4} \]
Substitute 4 for p
Why: Every p becomes 4, in brackets.
\[ h(4) = (4)^2 + 2(4) \]
Simplify
Why: Square, then double, then add.
\[ h(4) = 16 + 8 = 24 \]
Substitute 5 for m in the second rule
Why: Work inside the radical first.
\[ g(5) = \sqrt{5 - 4} = \sqrt{1} = 1 \]
Check both by reversing the arithmetic
Why: For h, 24 minus 8 is 16, which is 4 squared, as required. For g, 1 squared plus 4 is 5, the input we started with.
\[ 24 - 2(4) = 16 = 4^2, \qquad 1^2 + 4 = 5 \]
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, pp. 17-18 — Example 1.7 and Try It 1.4
These two are here for speed rather than difficulty, so try both before reading. What changes between them is the letter used for the input, and that change should mean nothing at all to you.
A warning about the second rule. The extracted text of this book loses radical signs, so the rule can look like a subtraction when it is actually a square root of a subtraction. It is the square root of the quantity m minus four, and the difference matters: evaluated at five it gives one either way, but at eight it gives two rather than four.
The check at the end is the habit to keep. Both of these rules can be undone: subtract and take a square root for the first, square and add for the second. Running the calculation backwards catches arithmetic slips in a couple of seconds and costs nothing.
Error analysis
Annotate
On: \( f(x) = x^2 - 4x: \quad f(-3) = -3^2 - 4\cdot -3 = -9 + 12 = 3 \)
Read the wrong solution first and try to locate the error before revealing the annotations. The arithmetic in it is all correct, which is what makes it hard to spot and dangerous to trust.
The error is in the very first substitution. The input is a negative number and it was written into the formula without brackets, so the square attached itself to the digit only and left the minus sign outside. That turns a positive nine into a negative nine, and the whole answer shifts by eighteen.
Notice that the reported answer looks perfectly reasonable. It is a small whole number, there is no division by zero, nothing is obviously broken. That is why the empty-slot habit matters: the error is invisible after the fact, so it has to be prevented rather than detected.
Worked example
The section's exercises ask for the same construction with a different rule. Try it before reading on.
\[ g(x) = 5 - x^2, \qquad \text{evaluate } \frac{g(x+h) - g(x)}{h}, \; h \ne 0 \]
Evaluate g at x plus h
Why: One copy of the input letter, inside a square, so one pair of brackets.
\[ g(x+h) = 5 - (x+h)^2 \]
Expand the square
Why: The minus sign is outside, so every term of the expansion changes sign.
\[ g(x+h) = 5 - x^2 - 2xh - h^2 \]
Subtract g of x
Why: The 5 and the minus x squared cancel against the second bracket.
\[ \frac{\left(5 - x^2 - 2xh - h^2\right) - \left(5 - x^2\right)}{h} = \frac{-2xh - h^2}{h} \]
Factor and cancel h
Why: Both surviving terms carry an h.
\[ \frac{h(-2x - h)}{h} = -2x - h \]
Check with numbers, taking x equal to 2 and h equal to 0.1
Why: The formula gives minus 4.1. Direct computation gives g of 2.1 equal to 0.59 and g of 2 equal to 1, and their difference over 0.1 is also minus 4.1.
\[ \frac{0.59 - 1}{0.1} = -4.1 = -2(2) - 0.1 \]
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, p. 35 — Exercise 32
Cover the steps and do this one yourself first. It is the same construction as Example 1.6 part d with two differences, and both of them trip people up.
The first difference is the minus sign sitting in front of the squared term. When you expand the square you must then change the sign of every term you produced, which is three sign changes in one step. Writing the expansion in brackets first and distributing the minus sign afterwards is slower and much safer.
The second difference is that the input letter is x rather than a, so the answer comes out in terms of x and h. That should feel like no difference at all. If it does feel like a difference, that is a sign the earlier example was being remembered rather than understood, which is worth knowing now rather than in an exam.
Section
Part 5
Concept
Two tasks use the same function and look almost identical on the page. Which one you are doing depends on which letter is known.
| you are given | you are looking for | the task | how many answers |
|---|---|---|---|
| the input | the output | evaluate | always exactly one |
| the output | the input | solve | none, one, or several |
\[ \text{evaluate: } f(2) = \;? \qquad \text{solve: } f(x) = 4 \]
The count in the last column is not a coincidence. Exactly one output per input is the definition itself, so evaluating can never return a list. Nothing forbids two inputs sharing an output, so solving can.
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, p. 16 — finding input and output values of a function
This table is the organising idea for the rest of the lesson, so take a minute over it. The two tasks use the same function and the same formula, and on a page they look nearly identical. What distinguishes them is which letter you were handed.
The last column is the part worth remembering, and it is not an accident. Evaluating can only ever produce one answer because exactly one output per input is the definition of a function. Solving has no such guarantee, because nothing forbids two inputs from sharing an output, so the honest expectation is a list.
From here on, the first thing to do with any question in this section is to decide which of the two it is. Almost every error in the second half of this lesson comes from doing one task when the other was asked for, or from expecting one answer when the question had two.
Worked example
An equation ties two letters together without saying which is the input. To write p as a function of n, get p alone on one side.
\[ 2n + 6p = 12 \]
Subtract 2n from both sides
Why: Move everything that is not the output to the other side.
\[ 6p = 12 - 2n \]
Divide both sides by 6
Why: Divide every term, not just the first.
\[ p = \frac{12 - 2n}{6} \]
Split the fraction and simplify
Why: 12 over 6 is 2, and 2 over 6 is one third.
\[ p = \frac{12}{6} - \frac{2n}{6} = 2 - \frac{1}{3}n \]
Name the rule
Why: Now p is written entirely in terms of n, so it is a function of n.
\[ p = f(n) = 2 - \tfrac13 n \]
Check by substituting the formula back into the original equation with n equal to 3
Why: The formula gives p equal to 1, and 2 times 3 plus 6 times 1 is 12, which is the equation we started from.
\[ f(3) = 2 - 1 = 1, \qquad 2(3) + 6(1) = 12 \;\checkmark \]
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, p. 19 — Example 1.9
An equation with two letters in it does not say which one is the input. Writing p as a function of n means choosing n as the input and then rearranging until p stands alone on one side with only n on the other.
The algebra is standard, but the division step is worth watching. When you divide both sides by six you must divide every term on the right, not just the first one. Splitting the fraction into two pieces, as the solution does, makes that impossible to forget.
The check at the end is the kind you should do automatically. Pick any input, run it through your new formula, and put both numbers back into the original equation. If the equation is satisfied, your rearrangement was sound. It takes ten seconds and catches almost every algebra slip.
Worked example
Not every equation can be rearranged into a function. Trying it is the fastest way to find out.
\[ x^2 + y^2 = 1 \]
Isolate y squared
Why: Subtract x squared from both sides.
\[ y^2 = 1 - x^2 \]
Take the square root of both sides
Why: A square root of a positive number has two signs, and both are genuine solutions.
\[ y = \pm\sqrt{1 - x^2} \]
Test one input
Why: Put x equal to 0.6 and both signs give a legal point on the circle.
\[ y = \pm\sqrt{1 - 0.36} = \pm 0.8 \]
Figure (svg): The unit circle drawn as two coloured arcs, the upper half in cyan and the lower half in red, with a dashed vertical line at x equals 0.6 meeting them at heights 0.8 and minus 0.8.
Now try Try It 1.6, which does work
Why: Here the power is odd, so the root has only one sign.
\[ x - 8y^3 = 0 \;\Longrightarrow\; y^3 = \frac{x}{8} \;\Longrightarrow\; y = \frac{\sqrt[3]{x}}{2} \]
Check both results at a sample input
Why: At x equal to 8 the cube root formula gives y equal to 1, and 8 minus 8 times 1 cubed is 0, so that one is a function. The circle gave two values of y at a single x, so it is not.
\[ y(8) = \frac{\sqrt[3]{8}}{2} = 1, \quad 8 - 8(1)^3 = 0 \;\checkmark \]
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, pp. 19-20 — Example 1.10 and Try It 1.6
The book puts these two next to each other for a reason, and the reason is the parity of the exponent. The same rearrangement works in one case and fails in the other, and the only difference is whether you are undoing a square or a cube.
Undoing a square introduces two signs, because both a positive and a negative number square to the same thing. So one input produces two outputs, and the relation cannot be a single function of x. Look at the picture: the upper arc and the lower arc are each perfectly good functions, but together they are not one.
Undoing a cube introduces only one sign, because cubing preserves sign. That is why the second rearrangement succeeds and produces an honest formula. The general lesson is that an even power in the equation is the warning sign, and it is the same warning sign you will meet again in the section on inverse functions.
Worked example
The output is given and the input is wanted, so this is an equation to solve, not a substitution to perform.
\[ h(p) = p^2 + 2p, \qquad \text{solve } h(p) = 3 \]
Replace h of p by its formula
Why: The equation now involves only p.
\[ p^2 + 2p = 3 \]
Subtract 3 from each side
Why: A quadratic must be set equal to zero before factoring.
\[ p^2 + 2p - 3 = 0 \]
Factor the quadratic
Why: Two numbers multiplying to minus 3 and adding to 2: they are 3 and minus 1.
\[ (p + 3)(p - 1) = 0 \]
Set each factor equal to zero
Why: A product is zero exactly when one of its factors is.
\[ p + 3 = 0 \;\Rightarrow\; p = -3, \qquad p - 1 = 0 \;\Rightarrow\; p = 1 \]
Figure (svg): The parabola h of p equals p squared plus 2p with its vertex at (-1, -1). A dashed horizontal line at height 3 meets it at (-3, 3) and (1, 3), and a separate dot marks (4, 24).
Check both answers in the original function
Why: Substituting minus 3 gives 9 minus 6, which is 3, and substituting 1 gives 1 plus 2, which is also 3. Both are genuine, and the graph shows the horizontal line at height 3 crossing twice.
\[ h(-3) = 9 - 6 = 3, \qquad h(1) = 1 + 2 = 3 \]
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, p. 18 — Example 1.8 and Figure 1.7
Notice what the question gave you. The output is known and the input is wanted, so this is an equation to solve, and you should already be expecting more than one answer before you start.
The algebra is the standard quadratic routine, and the step people skip is moving everything to one side first. Factoring only tells you something when the product equals zero, so the equation has to be arranged that way before the factoring is worth anything.
Look at the graph once you have both answers. The horizontal line at height three cuts the parabola twice, once on each side of the vertex, and those two crossings are your two solutions. The green dot is there as a contrast: it marks the evaluation at four, and an evaluation is always a single point.
Intuition
Evaluating walks up from the horizontal axis. Because the graph is a function, there is exactly one point above each input, so you can only ever land on one height.
Solving walks across from the vertical axis. Nothing in the definition limits how many points sit at a given height, so the walk can meet the curve any number of times: twice for the parabola at height 3, never at height minus 5.
\[ h(p) = -5: \quad p^2 + 2p + 5 = 0, \quad \text{no real solution} \]
So when you are asked to solve, finish the sweep. An answer of one value is only complete after you have checked that no other input reaches the same output.
This slide is short but it is the reason the previous three had the answers they did, so read it as an explanation rather than as a fact.
Going up from the horizontal axis you can only ever meet the curve once, because a function has exactly one output per input. That is not a feature of parabolas; it is true of every function you will ever draw, and it is the reason evaluating is safe.
Going across from the vertical axis you have no such protection. The curve may cross that height twice, or five times, or never. The parabola misses the height minus five entirely, so the corresponding equation has no real solution at all, and reporting no solutions is a complete and correct answer.
Worked example
Same two tasks again, now with only a graph to work from. The function is not given by a formula, and it does not need to be.
Evaluate f at 2
Why: Find x equal to 2 on the horizontal axis and climb to the curve.
\[ \text{the point is } (2, 1) \;\Longrightarrow\; f(2) = 1 \]
Figure (svg): The parabola with vertex at (1, 0) on a grid. A dashed line runs up from x equals 2 to the curve at the point (2, 1), then across to the vertical axis at height 1.
Now solve f of x equals 4
Why: This time start at 4 on the vertical axis and move horizontally, marking every crossing.
\[ \text{the crossings are } (-1, 4) \text{ and } (3, 4) \]
Read off the inputs
Why: The x-coordinate of each crossing is a solution.
\[ x = -1 \quad \text{or} \quad x = 3 \]
Figure (svg): The same parabola with a dashed horizontal line at height 4 cutting the curve at the points (-1, 4) and (3, 4), and dashed verticals dropping from each to the horizontal axis.
Check that both crossings really are on the curve
Why: The curve is the parabola with vertex at (1, 0), so the height at any x is x minus 1, all squared. At minus 1 that is minus 2 squared, which is 4, and at 3 it is 2 squared, which is also 4.
\[ f(-1) = (-2)^2 = 4, \qquad f(3) = (2)^2 = 4 \]
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, pp. 22-24 — Example 1.12 and Figures 1.8 to 1.10
The function here is given only by a picture, with no formula anywhere, and both tasks are still perfectly doable. That is worth noticing: nothing in the definition of a function requires a formula.
For the evaluation, go up from the input on the horizontal axis until you meet the curve, then across to read the height. The first figure shows that journey as two dashed segments. You will land on exactly one point, always.
For the equation, reverse the journey: start at the wanted output on the vertical axis and travel horizontally, marking every crossing. The second figure shows the line at height four cutting the curve twice, so there are two answers. The mistake to guard against is stopping at the first crossing you meet, and the cure is to carry the horizontal line all the way across the picture before you write anything down.
Prediction
\[ \text{using the same graph, solve } f(x) = 1 \]
Predict first
The curve is the parabola with vertex at the point (1, 0). How many solutions does the equation have, and what are they?
Correct: Two solutions, x equal to 0 and x equal to 2
Why: The horizontal line at height 1 cuts the parabola on both sides of its vertex, at the points (0, 1) and (2, 1). Reading only the right-hand crossing is the standard slip, and it is the same one the trap slide is about. The pair minus 1 and 3 belongs to the output 4, not 1.
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, p. 24 — Try It 1.8
Pick an option before revealing. Two of the four are designed to catch specific slips, so it is worth knowing which one tempted you.
The curve is the parabola with its lowest point at horizontal position one, so it is symmetric about that position. A horizontal line at height one therefore meets it at two places, one unit to each side of the vertex, which is at zero and at two.
One of the wrong options gives only the right-hand crossing, which is the stopping-too-early error again. Another gives the pair of answers that belongs to the output four, from the previous slide, which is a reminder that the answer depends on the height you started from and has to be recomputed each time.
Trap
An equation is solved and the first answer is reported:
\[ p^2 + 2p = 3 \]
\[ p = 1 \quad\text{(reported)} \]
Half an answer. The quadratic has two roots and only one was written down.
Factoring produces two factors, and each one gives a solution. Set both to zero before you stop.
\[ (p+3)(p-1) = 0 \]
\[ p = -3 \quad\text{or}\quad p = 1 \]
On a graph the same habit is a sweep: follow the horizontal line all the way across the picture, not just to the first crossing.
This is the most expensive habit in this part of the lesson, because the work is right and only the reporting is incomplete. A marker cannot give you credit for the solution you did not write.
The algebraic version has a mechanical cure. Factoring produces two brackets; set each one equal to zero and you will automatically write two lines. If you find yourself writing only one line after factoring, you have skipped a bracket.
The graphical version needs a different cure, and it is physical. Put your finger on the vertical axis at the wanted height and slide it all the way across the picture, to both edges, before you stop. Every crossing your finger passes is an answer.
Worked example
The section's exercises ask for both tasks side by side, which is exactly where the two get confused. Cover the solution and try it first.
\[ f(x) = x^2 - 3x: \quad \text{(a) evaluate } f(5), \quad \text{(b) solve } f(x) = 4 \]
Substitute 5 into the rule
Why: Input known, so this is evaluation.
\[ f(5) = (5)^2 - 3(5) = 25 - 15 = 10 \]
Set the rule equal to 4
Why: Output known, so this is an equation.
\[ x^2 - 3x = 4 \]
Move everything to one side
Why: Subtract 4 so the quadratic equals zero.
\[ x^2 - 3x - 4 = 0 \]
Factor
Why: Two numbers multiplying to minus 4 and adding to minus 3: minus 4 and 1.
\[ (x - 4)(x + 1) = 0 \]
Read both roots
Why: One from each factor.
\[ x = 4 \quad \text{or} \quad x = -1 \]
Check both roots in the original rule
Why: Substituting 4 gives 16 minus 12, which is 4, and substituting minus 1 gives 1 plus 3, which is also 4. Part (a) returned one number and part (b) returned two, which is what the two tasks should do.
\[ f(4) = 16 - 12 = 4, \qquad f(-1) = 1 + 3 = 4 \]
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, p. 36 — Exercise 37
The exercise sets in this book often put the two tasks together in parts a and b of one question, which is exactly the situation where they get confused. Try both parts before reading the steps.
Part a hands you the input, so it is a substitution and it produces a single number. Part b hands you the output, so it is an equation, and it goes through the same routine as Example 1.8: move everything to one side, factor, and set each factor to zero.
Compare the shapes of the two answers when you have finished. One number and two numbers, from the same rule, in the same question. If both of your answers came out as single numbers, you almost certainly solved part b by finding one root and stopping.
Section
Part 6
Concept
By convention the input goes on the horizontal axis and the output on the vertical one. The graph of f is then the set of all the points whose height is the output of their own horizontal position.
\[ \text{graph of } f = \{(x, y) : y = f(x)\} \]
So a point on the curve is a completed sentence. A point at horizontal position 0 and height 2 says that the output at 0 is 2, and nothing more.
\[ (0, 2) \text{ on the graph} \;\Longleftrightarrow\; f(0) = 2 \]
If a function is defined at only a few inputs, its graph is only a few dots. A continuous curve just means there are infinitely many of them.
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, pp. 24-26 — graphs of functions and Figure 1.11
A graph looks like a picture and behaves like a list. Every point on it is one completed statement about the function, and the curve is simply all those statements drawn at once.
Practise reading a point as a sentence. A point at horizontal position zero and height two says that the output at zero is two. That is the entire content of the point. Being able to go back and forth between a dot and an equation is what makes the two line tests obvious rather than magical.
The convention that inputs go across and outputs go up is arbitrary but universal, and both of the tests you are about to meet depend on it completely. If a question ever plots things the other way round, the tests swap over with it.
Picture it
Figure (svg): Two panels. On the left a rising and falling cubic with a dashed vertical line crossing it once. On the right a parabola opening to the right with a dashed vertical line crossing it twice.
A vertical line is one fixed input. Every point where it meets the picture is an output that input produces, so two meetings means two outputs, which is exactly what the definition forbids.
On the right the picture is a perfectly good relation, the set of points whose horizontal coordinate is the square of their height. It is just not a function of x, because the input 2 carries two different heights.
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, p. 26 — the vertical line test and Figure 1.12
The vertical line test is usually taught as a rule to memorise. It is better understood as a translation, so spend a moment connecting the picture to the words.
A vertical line is a single input: every point on it has the same horizontal position. So the places where that line meets the picture are exactly the outputs that input produces. One meeting means one output, which the definition wants. Two meetings means two outputs, which the definition forbids.
The right-hand picture is not a broken drawing. It is the set of points whose horizontal coordinate is the square of their height, which is a perfectly respectable relation and comes up in the chapter on conic sections. It is simply not a function of the horizontal variable, because the input two carries two heights.
Intuition
The test is not a new fact to memorise. It is the definition translated into geometry, one word at a time.
| in the definition | on the graph |
|---|---|
| one input value | one vertical line |
| the outputs it produces | the points where that line meets the curve |
| exactly one output | exactly one meeting point |
| two outputs, so not a function | two meeting points, so not a function |
This is worth doing deliberately, because the horizontal line test is about to be built the same way from a different sentence, and telling the two apart is much easier if you know which sentence each came from.
This table is here so that you never have to memorise the vertical line test as a separate fact. Each row translates one phrase of the definition into one feature of the picture, and the test falls out of the translation.
Doing the translation deliberately has a practical payoff two slides from now. The horizontal line test is built from a different sentence in exactly the same way, and if you know which sentence each test came from, you will never mix up which conclusion you are entitled to draw.
It is also useful when a question is phrased oddly. If you are ever unsure what a test is telling you, translate the picture back into the language of inputs and outputs and reread the definition. The definition is short, and it always settles the matter.
Worked example
Three pictures. For each one, ask whether any vertical line meets it more than once.
Figure (svg): Three panels: (a) a curve that rises to a peak near x equals minus 1, dips to a trough near x equals 1.5 and rises again; (b) a straight line falling from left to right; (c) a circle of radius about 2.7 with a dashed vertical line at x equals 1.8 cutting it at two points.
Sweep across the first graph
Why: It rises, falls and rises again, but at every horizontal position there is only one point of the curve.
\[ \text{(a) a function} \]
Sweep across the second graph
Why: A straight line with negative slope: one point above each input.
\[ \text{(b) a function} \]
Sweep across the third graph
Why: A vertical line at 1.8 cuts the circle at two points, one above the axis and one below.
\[ \text{(c) not a function} \]
Say where the third one fails
Why: Every vertical line strictly between the left and right edges of the circle meets it twice.
\[ \text{fails for every } x \text{ with } -2.7 < x < 2.7 \]
Check the circle against the algebra
Why: A circle of radius 2.7 has the equation x squared plus y squared equals 7.29, and solving that for y gives plus or minus the square root, two values, which is the same failure seen in Example 1.10.
\[ y = \pm\sqrt{7.29 - x^2} \]
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, pp. 26-28 — Example 1.14 and Figures 1.13, 1.14
Three pictures, one question asked three times. Resist the urge to judge them by how complicated they look; the wiggles in the first one are completely irrelevant to the test.
The first two pass. Put your finger anywhere on the horizontal axis, move straight up, and you meet the curve once. The third fails, and the dashed line shows one place where it fails, at horizontal position one point eight, where the line enters the circle at the top and leaves it at the bottom.
The last two steps are worth doing even though the answer is already clear. Saying where the test fails, rather than just that it fails, is what a full answer looks like, and the algebra confirms it: solving the circle's equation for the height produces a plus-or-minus sign, which is the same two-outputs failure you saw in Example 1.10 in a different costume.
Concept
Being a function controls what leaves each input. A second, stricter condition controls what arrives at each output.
one-to-one function — A function in which each output value corresponds to exactly one input value.
\[ f(a) = f(b) \;\Longrightarrow\; a = b \]
A school that awards only the letter grades A, B, C and D, worth 4.0, 3.0, 2.0 and 1.0, has a one-to-one rule: each letter gives one average, and each average comes from one letter.
| letter grade | grade point average |
|---|---|
| A | 4.0 |
| B | 3.0 |
| C | 2.0 |
| D | 1.0 |
Squaring is the standard function that is not one-to-one, because minus 3 and 3 both arrive at 9. Notice it is still a function; one-to-one is an extra demand on top.
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, pp. 24-25 — one-to-one functions and Table 1.13
Here is the extra condition that was promised several slides ago. Being a function controls what leaves each input. Being one-to-one controls what arrives at each output, and it is an additional demand, not a replacement.
Read the symbolic line as a sentence: if two inputs produce the same output, then they were not really two inputs at all, because they must be equal. That is the cleanest way to state one-to-one, and it is the form you will actually use when proving it.
The letter-grade table is one-to-one and the percent-grade table from Example 1.2 is not, even though both are functions. The difference is that each letter grade corresponds to its own average, with nothing shared, while a whole band of percents shared an average. Squaring is the standard function that fails: it is a function, always was, and is simply not one-to-one.
Worked example
Two questions in one, and they must be answered in order: first a function, then one-to-one.
\[ A = \pi r^2, \qquad r > 0 \]
Check that area is a function of radius
Why: Each radius gives one number when you square it and multiply by pi.
\[ r = 3 \;\Longrightarrow\; A = 9\pi \approx 28.27 \]
Now ask the reverse question
Why: Given an area, is the radius forced? Solve the formula for r.
\[ r^2 = \frac{A}{\pi} \;\Longrightarrow\; r = \pm\sqrt{\frac{A}{\pi}} \]
Use the fact that a radius is positive
Why: The negative root is not a radius, so only one value survives.
\[ r = \sqrt{\frac{A}{\pi}} \]
Figure (svg): Two panels. On the left the parabola y equals x squared with a dashed horizontal line at height 4 meeting it at x equals minus 2 and x equals 2. On the right the curve A equals pi r squared drawn only for positive r, with a dashed horizontal line at about 28.3 meeting it once, at r equals 3.
Check the two panels against each other
Why: Squaring every real number sends both minus 2 and 2 to 4, so a horizontal line hits twice. Restricting the input to positive radii deletes the left branch, and the same horizontal line now hits once, at a radius of 3 for an area of about 28.27.
\[ \sqrt{\frac{9\pi}{\pi}} = \sqrt{9} = 3 \;\checkmark \]
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, p. 25 — Example 1.13
Two questions, and the order is not optional. Being one-to-one is a property of functions, so the first thing to establish is that you have a function at all.
The first part is quick: every radius gives one area, because squaring and multiplying are single-valued operations. The second part is where the thinking happens. Solve the formula for the radius and a plus-or-minus sign appears, exactly as it did for the circle, and for a moment it looks as though this will fail too.
What rescues it is the domain. A radius is a positive length, so the negative root is not a legal answer and only one value survives. Look at the two panels: the left one allows every real input and a horizontal line hits twice, the right one allows only positive inputs and the left branch is simply gone. Restricting the domain to repair the horizontal line test is a move you will use again when you build inverse functions.
Concept
Once a graph has passed the vertical line test, one more sweep decides whether it is one-to-one, and it is the same idea turned through a right angle.
| sweep | the line stands for | more than one hit means |
|---|---|---|
| vertical | one input | that input has two outputs, so not a function |
| horizontal | one output | that output has two inputs, so not one-to-one |
\[ \text{one horizontal line, two hits} \;\Longrightarrow\; \text{not one-to-one} \]
The order matters. A picture that fails the vertical test is not a function at all, and asking whether it is one-to-one is asking a question about something that does not exist.
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, p. 28 — the horizontal line test
The second test is the first one turned through a right angle, and the table says exactly what changes. A vertical line is one input, a horizontal line is one output, and the conclusion you are allowed to draw changes with it.
Be precise about what a failure means. Two crossings of a horizontal line tell you that one output is shared by two inputs. That breaks one-to-one and nothing else. It says absolutely nothing about whether the picture is a function, and the trap slide coming up is about exactly that confusion.
The ordering rule at the bottom is worth taking seriously. If a picture fails the vertical line test there is no function there, and asking whether it is one-to-one is asking a question about an object that does not exist. Always run the vertical sweep first.
Worked example
Take the two pictures from Example 1.14 that did pass the vertical line test and sweep them again, this time horizontally.
Figure (svg): Two panels. On the left the wiggly curve with a dashed horizontal line at height 1 crossing it at three points. On the right a straight falling line with a dashed horizontal line crossing it at one point.
Sweep the first graph
Why: The line at height 1 meets the curve three times, near minus 1.76, near zero and near 2.54.
\[ f(-1.76) = f(-0.03) = f(2.54) = 1 \]
Conclude for the first graph
Why: Three different inputs sharing one output.
\[ \text{(a) a function, but not one-to-one} \]
Sweep the second graph
Why: A straight line with nonzero slope meets any horizontal line exactly once.
\[ \text{(b) one-to-one} \]
Say why the straight line can never fail
Why: If two inputs gave the same output, the segment joining those two points would be horizontal, and a line of slope minus one half has no horizontal piece.
\[ y = 2 - \tfrac12 x: \quad \text{slope} = -\tfrac12 \ne 0 \]
Check the algebra behind that claim
Why: Setting the outputs at a and b equal gives 2 minus half a equals 2 minus half b, and cancelling leaves a equal to b, which is exactly the one-to-one condition.
\[ 2 - \tfrac12 a = 2 - \tfrac12 b \;\Longrightarrow\; a = b \]
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, pp. 28-29 — Example 1.15 and Figure 1.16
Notice which two pictures were chosen. These are the two from Example 1.14 that passed the vertical line test, because those are the only ones for which this question makes sense.
The first one fails, and it fails badly: the line at height one crosses three times, so three different inputs all produce the same output. Any curve that goes up, comes down and goes up again will fail this test somewhere, which means most of the interesting curves in this course are not one-to-one.
The second one passes, and the last two steps explain why it could never have failed. A straight line with nonzero slope has no horizontal piece anywhere, so two points of it cannot share a height. The algebraic version of that argument is the one to copy: assume two inputs give the same output, do the algebra, and watch the two inputs turn out to be equal.
Trap
A horizontal line cuts the parabola twice, and the conclusion comes out wrong:
\[ f(x) = x^2: \quad f(-2) = f(2) = 4 \]
\[ \Longrightarrow\; \text{not a function?} \]
Wrong. Squaring is a function, and always was.
Two inputs sharing an output breaks one-to-one, not the definition of a function. The definition is only broken when one input has two outputs.
\[ \text{not one-to-one} \ne \text{not a function} \]
The test you used tells you which conclusion you are entitled to. A horizontal line never says anything about whether the picture is a function.
These two conclusions get swapped constantly, and the reason is that both tests look like a line crossing a curve twice. What differs is which line, and therefore which sentence you are testing.
Say it once more in full. A vertical line crossing twice means one input, two outputs, so not a function. A horizontal line crossing twice means two inputs, one output, so not one-to-one but still a function. The parabola is the example that settles it: it fails the horizontal test and it is unquestionably a function.
A practical rule: before you write a conclusion, name which line you drew. If you drew a horizontal line, the word function cannot appear in your conclusion at all, only the words one-to-one.
Tweak it
Parameter explorer
This is a cubic with one adjustable coefficient. Slide it and watch the shape. For which values of the parameter does a horizontal line ever meet the curve more than once?
\[ f(x) = x^3 + p\,x, \qquad p = {p} \]
Start with the coefficient at zero, which gives the plain cubing rule, and notice the shape: it only ever rises. A curve that only rises can never meet a horizontal line twice, so it is one-to-one.
Now push the coefficient up into the positive numbers. The curve gets steeper but it still only rises, so nothing changes. This is the half of the family that is one-to-one, and it is the boring half.
Then pull the coefficient below zero and watch a hump and a dip appear. As soon as they exist, any horizontal line drawn between them meets the curve three times, and the function stops being one-to-one. Try to find the exact coefficient where the behaviour switches, and notice that it is the value at which the hump and the dip collapse into each other.
Counterexample
\[ f(x) = 3(x - 5)^2 + 7 \]
Discussion prompt
Show that this function is not one-to-one. A single well-chosen pair of numbers is a complete proof, so do not draw anything unless you want to.
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, p. 39 — Exercise 92
Write your answer before revealing, and notice how little you need. Disproving a general claim takes one counterexample, so a single well-chosen pair of numbers is a complete and rigorous proof.
Choosing the pair is the only skill involved. The rule squares a bracket, and squaring treats numbers symmetrically about the point where the bracket is zero, which here is five. So pick any two inputs placed equally either side of five and they are guaranteed to give the same output.
Compare this with what it would take to prove that a function is one-to-one. There you cannot check examples, because you would have to check infinitely many; you would need the algebraic argument used for the straight line. Disproving is cheap, proving is not, and knowing which one a question is asking for saves a lot of wasted effort.
Concept
The section closes with a set of building blocks. Learn these nine shapes by name now, because every graph in the rest of the course is one of them stretched, shifted or reflected.
Figure (svg): Nine small graphs on one sheet: a horizontal line, a diagonal line, a V shape, a parabola, an S-shaped cubic, a two-branch hyperbola, a pair of branches both above the axis, a half parabola lying on its side starting at the origin, and a flat S-shaped cube root curve.
All nine pass the vertical line test, since each is defined by a formula that produces one output. Only four pass the horizontal line test: the identity, the cubic, the square root and the cube root. The reciprocal passes on each branch but fails as drawn only if you forget that it never repeats a value.
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, pp. 29-32 — Table 1.14, the toolkit functions
These nine are worth learning by heart, by name, shape and formula. Chapter 1 spends the rest of its sections transforming them, and every graph you sketch for the next several months will be one of these shifted, stretched or flipped.
Use the sheet to practise both tests at once. Every one of the nine passes the vertical line test, because each is given by a formula that produces a single output. Only some pass the horizontal line test, and you can see which at a glance: the ones that rise all the way across, or fall all the way across, are the one-to-one ones.
Two of them deserve a second look. The absolute value and the quadratic both fold the input line in half, so both fail the horizontal test in the obvious way. The two reciprocal shapes behave differently from each other: one falls on both branches and never repeats a value, while the other is symmetric and takes each positive value twice.
Comparison
Comparison matrix
| picture | passes the vertical test | passes the horizontal test |
|---|---|---|
| the line y = 2 − x/2 | yes | yes |
| the parabola y = x² | yes | no |
| the circle x² + y² = 9 | no | no |
| the sideways parabola x = y² | no | yes |
| the cubic y = x³ | yes | yes |
Fill in every blank before checking. Each row is a picture you have already met in this lesson, so look back rather than guessing if one of them is not coming to mind.
The two circle-like rows are the ones to be careful with. Both fail the vertical test, which means neither is a function, and once that is established the horizontal column is not really a meaningful question. The sideways parabola is listed as passing the horizontal test because no output value repeats on it, which is true and also completely useless, since it is not a function in the first place.
The cubic row is the only one where both answers are yes, and it is worth knowing why. The cubing rule rises steadily across the whole picture, so it meets every vertical line once and every horizontal line once, which is exactly what makes it invertible later in the course.
Section
Part 7
Pattern
Figure (svg): A flow diagram. A relation leads to the question whether one input ever gives two outputs; yes leads to not a function, no leads to a function. A function then leads to the question whether one output ever comes from two inputs; yes leads to not one-to-one, no leads to one-to-one.
This is the order to work in for every question in this section. It is written as a flow because the questions depend on each other: the second one is only worth asking once the first has been answered.
The first bullet is the one most often skipped, and it is the one that caused the different answers in the menu example and the grade example. Before anything else, name which quantity is the input. Swapping the two can change a yes into a no.
The fourth bullet is the habit that carries through the rest of the chapter. Look at which letter you were handed. An input means substitute and expect one number. An output means solve and expect a list. And the last bullet is the conclusion discipline: the test you ran determines which word you are allowed to use in your answer.
Ranking
Put in order
Order the steps of a complete check that a graph shows a one-to-one function.
Why: The direction has to be settled first, because the whole test depends on which quantity is the input. The vertical sweep comes before the horizontal one: if the picture is not a function at all, asking whether it is one-to-one is asking about something that does not exist.
Drag all five into order before checking. Only one of the orderings is defensible, and the reason is logical rather than stylistic.
The first step is the one that feels optional and is not. Everything downstream depends on which axis carries the input, so settling that has to come before either sweep.
The middle step, concluding that the picture is a function, looks like a formality that could be left out. It cannot. It is what licenses the second question, because one-to-one is a property that only functions can have. Writing it down also forces you to notice when the vertical sweep has failed and the rest of the question is moot.
Check
Check your understanding
Which of these relations is a function of x?
Answer: C
Why: In option C the inputs 1, 2 and 3 are all different, so each has exactly one output. The output 7 repeating three times is allowed, because the definition only ever looks at the inputs.
Three of the four options fail for the same reason, so the useful thing is to name the repeated input in each one before you choose.
The correct option has three distinct inputs and one output repeated three times, which is the pattern the whole first part of this lesson was about. If a repeated output still makes you hesitate, go back to the mapping diagrams: several arrows arriving at one place is allowed, several arrows leaving one place is not.
The two equations among the options are both failures of the same kind, and both can be settled either by solving for the output, which produces a plus-or-minus sign, or by imagining a vertical line, which cuts each picture twice.
Check
Check your understanding
For f(x) = x² − 3x, which statement is correct?
Answer: B
Why: Evaluating at 5 gives 25 minus 15, which is 10, and it is a single number because 5 is a single input. Solving gives the quadratic x squared minus 3x minus 4 equals 0, which factors into x minus 4 times x plus 1, so the inputs 4 and minus 1 both produce the output 4.
Each option makes two claims, so check both before committing. It is quite possible to get the evaluation right and still pick the wrong option.
The evaluation is a single substitution, and the trap in one of the options is forgetting the second term entirely and reporting just the square. Work the substitution with the input in brackets and both terms will survive.
The equation part is the quadratic routine again. Move everything to one side, factor, and set each bracket to zero, which gives two distinct inputs. One of the options claims that the evaluation itself has two values, which is impossible for any function at all, and recognising that immediately is a good sign that the definition has landed.
Check
Check your understanding
A graph passes the vertical line test but fails the horizontal line test. What follows?
Answer: C
Why: Passing the vertical test is exactly what it means to be a function: no input has two outputs. Failing the horizontal test means some output is shared by two inputs, which breaks one-to-one and nothing else. The parabola is the standard example.
This question is purely about which conclusion each test entitles you to, so answer it by naming what each line stands for rather than by picturing a particular curve.
Passing the vertical test is precisely what being a function means, so the first half of the answer is already settled. Failing the horizontal test means that some output is shared by two inputs, and that breaks one-to-one and nothing else.
The parabola is the picture to keep in mind here, since it does exactly this: it passes the first test, fails the second, and is a function that is not one-to-one. That combination is the most common one you will meet all year.
Explain it to yourself
Discussion prompt
Without looking back, write one sentence saying what a vertical line stands for and what two crossings mean, and a second sentence doing the same for a horizontal line. Then say which test has to be run first, and why.
Write your two sentences out before revealing. Being able to state these from memory, with the reason attached, is worth far more than being able to apply them, because the application is easy and the confusion between them is what costs marks.
The structure of each sentence should be the same: say what the line stands for, then say what two crossings therefore mean. If your sentences do not contain the words input and output, they are describing the picture rather than explaining it, and they will not protect you when the two tests get mixed up.
The third part, about the order, is the real test of understanding. A picture that fails the vertical sweep is not a function, so there is no function there to be one-to-one or otherwise. That is why the sweeps have to happen in that order and not the other.
Real world
Discussion prompt
A city's weekly garbage G, in tons, is a function of its population p, in thousands, written G = f(p). The town of Tola has 40,000 people and produces 13 tons a week. Write that in function notation, then say in plain English what f(5) = 2 means.
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, pp. 39-40 — Exercises 85 and 87
Write both answers before revealing, and be careful with the units, because that is where this kind of question is actually difficult.
The population is measured in thousands, so a town of forty thousand people is an input of forty. Writing the input as forty thousand is the standard error, and it produces a statement that is off by a factor of a thousand while looking perfectly sensible. Read the definition of the input variable before you write anything.
The second half of the slide is the reading habit that makes word problems tractable. Any statement of the form f of something equals something else is a complete sentence once you know what the input and output mean. And notice the last remark: asking in which years the duck count was thirty is a solving question, so the honest expectation is a list of years, not a single one.
Exit ticket
\[ f(x) = 8 - 3x, \qquad g(m) = \sqrt{m-4} \]
Discussion prompt
Three things, in order. Evaluate f at minus 2. Solve the equation f of x equals minus 1. Then say whether f is one-to-one, and give the reason a horizontal line would give.
OpenStax, Precalculus, §1.1 Functions and Function Notation §1.1, p. 35 — Exercise 35
Three tasks that together use most of the lesson. Do all three before revealing, and write the answers out properly rather than in your head.
The first two are the evaluate-and-solve pair one last time, on a rule simple enough that any error is an arithmetic one. Watch the negative input in the first: the rule subtracts three times the input, and subtracting three times a negative number adds.
The third is the one that shows whether the last part of the lesson landed. The graph is a straight line with nonzero slope, so a horizontal line meets it exactly once no matter what height you choose, and the function is one-to-one. Notice that this time solving returned a single answer, and notice why: it is one-to-one, which is exactly the condition that guarantees at most one input per output.
Recap
| question | what you look at | what a yes means |
|---|---|---|
| is it a function? | the inputs only, or a vertical line | each input has exactly one output |
| what is f(3)? | the input 3 | evaluate: substitute, and expect one number |
| solve f(x) = 3 | the output 3 | solve: expect none, one, or a list |
| is it one-to-one? | the outputs, or a horizontal line | each output comes from exactly one input |
\[ f(a+h) \ne f(a) + f(h), \qquad \text{the brackets are a slot, not a product} \]
Section 1.2 takes the definition further by asking which inputs are legal in the first place, which is the domain, and which outputs actually occur, which is the range.
OpenStax Algebra and Trigonometry 2e, §3.1 Functions and Function Notation §3.1 — the same material in Algebra and Trigonometry
Four questions, and the first thing to do with any problem in this section is to work out which of the four it is. The left column is what you are asked, the middle column is where to look, and confusing the middle column is where nearly all the errors come from.
The two rows in the middle are the pair that gets mixed up. If the input is given, substitute and expect exactly one number, because that is the definition of a function. If the output is given, solve and expect a list, because nothing forbids two inputs from sharing an output.
The line at the bottom is the notation habit to carry forward: the brackets after a function name mark a slot, not a multiplication, so you cannot distribute across them. Section 1.2 picks up where the definition left off, asking which inputs are legal in the first place and which outputs actually occur, which are the domain and the range.
Want this taught 1-on-1? Alexander tutors Precalculus — $55/session, free consultation.