Week 2: Monday

The Intermediate Value Theorem

A key property of continuous functions: If they hit two different y-values, they have to hit
everything in between. This makes sense from the informal definition of continuity: if you
can draw the graph from f (a) to f (b) without picking up the pencil, you have to cross the
space between them.

Example: the weather today. It started out cool (60 degrees) in Westwood this morning,
but as the sun rose, it got hotter, up to a high of 78 degrees in the afternoon. So there must
have been some time today where it was exactly 72 degrees.

Let’s make this formal:

The Intermediate Value Theorem: Suppose f (x) is continuous on the interval [a, b].
Then for any value M , if f (a) < M < f (b), then there is some c with a < c < b so that
f (c) = y.

When you see a theorem in mathematics, it is often useful to check whether or not the
assumptions are actually necessary. Here, they are. Consider the following function:
(
x x<0
f (x) =
x+1 x≥0

−4 −2 2 4

−2

−4

Note that f (−1) = −1; f (1) = 2. But there is no x-value where f (x) = 1/2, because it
jumps straight from 0 to 1.

[Here I told a terrible joke, that I will not repeat here. You’re welcome.]

So what can we do with this? This is a useful tool to show that equations can be solved.

Example: Show that sin(x) = 0.2 has a solution in the interval [0, π/2]:

1
sin(0) = 0. sin(π/2) = 1. Since 0.2 is between 0 and 1, there must be c between 0 and π/2
such that sin(c) = 0.2.

Zeros of functions
One way to use IVT is to prove that a function has a zero in a specific interval.

Consequence of IVT: Suppose f is a continuous function, and f (a) < 0 and f (b) > 0.
Then there is some c between a and b so that f (c) = 0.

Example:

f (x) = −x3 + 5x2 − 2. When x = 0, f (x) = −2. When x = 1, f (x) = 2. So IVT tells us
that there must be a root between 0 and 1.

Example: Show that 4 sin(x) = x + 1 has a solution in the interval (0, π/2).

Write as a single function: f (x) = 4 sin(x) − (x + 1) = 4 sin(x) − x − 1.

At 0: f (x) = −1.

At π/2 : f (x) = 4(1) − π/2 − 1. π/2 is less than 2, so this is positive.

So by IVT, there is some c in the interval (0, π/2) so that f (c) = 0. So in particular,
4 sin(c) = x + 1.

Important: IVT just tells us that there is a root, not what it is.

But: we can use it to approximate roots.

Technique: Find an interval where a root must exist. Split it into two pieces; a root must
be in one of them. Repeat.

Example:

Show that f (x) = cos2 (x) − 2 sin(x/4) has a root in (0, 2). Then estimate the root within
1/16.

Note that f is continuous. f (0) = 1; f (2) is around −0.786. So by IVT, there must be a
root.

Now, we look at (0, 1) and (1, 2). To figure out what’s happening, we look at the midpoint:
f (1) is around −0.203. f (0) is positive, f (1) is negative, so there is a root between 0 and 1.

Lets do this again, looking at 1/2: f (1/2) ≈ 0.521

So the zero is in (1/2, 1).

2
Again: f (3/4) ≈ 0.163. So the zero is in (3/4, 1).

The size of this interval is 1/4. So if we guess the midpoint for our zero, we will be at most
1/8 from the actual zero.

A Fun Extra: Temperature

Take a map, draw a circle on it.

Claim: there is a pair of points opposite each other on the circle, that have the same
temperature.

Proof:

Let T (θ) be the temperature of the point at angle θ.

Let f (θ) = T (θ) − T (θ + π). A zero of f means the temperature is equal at θ and the point
opposite it.

f (0) = T (0) − T (π).

f (π) = T (π) − T (π + π) = T (π) − T (0) = −f (0).

So either both are zero, and we are done; or they are opposite signs. Then by IVT, there
is some angle c such that f (c) = 0. This means that T (c) = T (c + π), so c and c + π give
opposite points with the same temperature.

3
Wednesday
The Derivative
The goal: find the slope of a tangent line to a function f at a point a. This can be thought
of as the "instantaneous rate of change."

15

10

−1 1 2 3 4

How do we do this? To find a slope we need two points, after all. So we pick a second point
that is near a, and find the slope of the line through those:

15

10

−2 −1 1 2 3 4

−5

Call the second point x. What is the slope here? Change in y over change in x. So we get:

f (x) − f (a)
x−a

This expression is called the difference quotient.

As we move x closer to a, we get a better approximation.

4
 15

10

−2 −1 1 2 3 4

−5

Theme of the course: find better and better approximations, take the limit.

The derivative of f (x) at a is defined as:

f (x) − f (a)
f ′ (a) = lim
x→a x−a

If this limit exists, we say that f is differentiable

This is the slope of the tangent line at a.

Example:

f (x) = x2 .

f (x) − f (0)
f ′ (0) = lim
x→0 x−0
x −0
2
= lim
x→0 x
= lim x
x→0
=0

Notice that the limit in the first stage was an indeterminate form: we had to simplify the
numerator and cancel. This will always be the case when finding derivatives: the top and
the bottom will both go to zero.

5
What about at a different point?
f (x) − f (3)
f ′ (3) = lim
x→3 x−3
x2 − 9
= lim
x→3 x − 3
(x + 3)(x − 3)
= lim
x→3 x−3
= lim x + 3
x→3
=6

So this is the idea: find f (a), plug it into the difference quotient, simplify, and then take the
limit.

Here’s another way of writing the derivative at a. Rather than focusing on the point x, we
can look at the distance x is from a, and take the limit as that distance goes to zero. We call
this distance h. We get the derivative by swapping x with a + h in the difference quotient:

f (a + h) − f (a) f (a + h) − f (a)
f ′ (a) = =
(a + h) − a h

Both of these will be useful. The second one is valuable because it lets us vary what a is
without changing the calculations very much.

Lets find the derivative of x2 at 3 again with this new format:

(3 + h)2 − 32
f ′ (3) = lim
h→0 h
h2 + 6h + 9 − 9
= lim
h→0 h
h(h + 6)
= lim
h→0 h
= lim h + 6
h→0
=6

The key here is to be careful with f (x + h). You get this by substituting x + h in for x; this
is not the same as f (x) + h. This is an incredibly common mistake, so watch out for it.

Armed with the derivative, we can write down an equation for the tangent line itself. Re-
member point-slope form: the equation for a line with slope m passing through (x0 , y0 )
is
(y − y0 ) = m(x − x0 )

6
In our case, the point is (a, f (a)), and the slope is the derivative. So the equation for the
line tangent to f at a is:
(y − f (a)) = f ′ (a)(x − a)

Derivatives of Linear Functions

The derivative is the slope of the tangent line. So if the function itself is a line, the derivative
should just have the same slope. Lets check that. We’ll start with horizontal lines:

Derivative of f (x) = c, at a point a:

f (a − h) − f (a)
f ′ (a) = lim
h→0 h
0−0
= lim
h→0 h
= lim 0
h→0
=0

Derivative of f (x) = mx + b, at a point a:

f (x + h) − f (x)
f ′ (a) = lim
h→0 h
m(x + h) + b − (mx + b)
= lim
h→0 h
mx + mh + b − mx − b
= lim
h→0 h
mh
= lim
h→0 h
mh
= lim
h→0 h
= lim m
h→0
=m

Differentiability
When might f not be differentiable? In general, if f ′ (a) exists, then f must be continuous
at a. But there are continuous functions that are not differentiable:

Example: f (x) = |x| is not differentiable at 0.

7
Let’s look at the right limit. When x > 0, |x| = x.

|0 + h| − |0|
lim+
h→0 h
Since h is positive, |0 + h| = h.
h
= lim+ = 1.
h→0 h

On the other hand, when h < 0, |0 + h| = −h. So we get

|0 + h| − |0| −h
lim− = lim− = −1.
h→0 h x→0 h

Since these don’t line up, the limit doesn’t exist. In general, if this happens, we get either
a corner or a cusp. A corner is what it sounds like; a cusp is when you have two smooth
curves that meet at the same point, but their tangent lines don’t match up.

A corner:

1.5

0.5

−2 −1 1 2

A cusp:

8
 2

1.5

0.5

−2 −1 1 2

You can also have a situation where the derivative isn’t defined because the tangent line is
vertical:

0.5

−2 −1 1 2
−0.5

−1

9
Friday
Derivative as a Function
Last time, we talked about the derivative at a point:

f (a + h) − f (a)
f ′ (a) = lim
h→0 h

It is often more useful to think of the derivative as a function; it takes in a point x, and spits
out the slope at that point. So we would write it as:

f (x + h) − f (x)
f ′ (x) = lim
h→0 h

The domain of f ′ (x) is all points where the limit exists.

Some examples:

Derivative of 3x2 − 2

Notation:

Another notation for the derivative: dy

dx
.

Spoken as "the derivative of y with respect to x."

For instance if y = 3x2 − 2, we would write dy

dx
= 6x.

We could also write: d

dx
3x2 − 2 = 6x.

Derivative Rules
From last time:
d
dx
x = 1, d
dx
mx = x, d
dx
c = 0.

The Power Rule: d n

dx
x = nxn−1 .

In words: bring the power down as a coefficient, then subtract off one in the exponent.

Let’s see why this is true.

d n (x + h)n − xn
x = lim
dx h→0 h

10
If you expand (x + h)n , you get xn + nxn−1 h + [terms that have larger powers of h]

xn + nxn−1 h + [terms with a factor of h2 ] − xn

= lim
h→0 h
nxn−1 h + [terms with a factor of h2 ]
= lim
h→0 h
cancel an h

= lim nxn−1 + [terms with a factor of h]

h→0
= nx n−1

Our proof above is just for natural numbers. It is also true for all exponents: natural
numbers, fractions, etc. More on that later.

Sums:If f and g are differentiable, then (f +g) is differentiable, and (f +g)′ (x) = f ′ (x)+g ′ (x).

Let’s show this:

(f + g)(x + h) − (f + g)(x)
(f + g)′ (x) = lim
h→0 h
f (x + h) + g(x + h) − f (x) − g(x)
= lim
h→0 h
(f (x + h) − f (x)) + g(x + h) − g(x)
= lim
h→0 h
Remember the limit laws: limit of sum is the sum of the limits.
f (x + h) − f (x) g(x + h) − g(x)
= lim + lim
h→0 h h→0 h
= f (x) + g (x)
′ ′

Constant Multiples: If f is differentiable, (cf )′ (x) = c · f ′ (x).

cf (x + h) − cf (x)
(cf )′ (x) = lim
h→0 h
factor out the c...
c(f (x + h) − f (x))
= lim
h→0 h

11
by the limit laws, we can pull the constant multiple outside of the limit
f (x + h) − f (x)
= c · lim
h→0 h
= c · f (x).
′

If we put this together, this lets us quickly find the derivative of any polynomial. Polynomials
are just sums of powers with constant multiples, and we have tools to handle all of that.

Example: d
dx
3x1 4 − 5x6 + 17x2 + 8.

We could, theoretically, put this into the definition of the derivative, work out all the details,
and calculate this thing. But we don’t have to. This is a sum of a bunch of terms; the Sum
Rule says we can compute the derivatives of each individually and add them up at the end.
So lets do that.

• d
dx
3x14 = 3 dx x = 3 · (14)x13 = 42x1 3
d 14

• d
dx
− 5x6 = −x dx x = −5 · 6x5 = −30x5
d 6

• d
dx
17x2 = 17 · d 2
dx
x = 17 · 2x = 34x

• d
dx
8 = 0.

Putting it all together, we get: d

dx
3x14 − 5x6 + 17x2 + 8 = 42x1 3 − 30x5 + 34x.

Example: d 4
dx
x − 3x2 + 2 = 4x3 − 6x.
d √
Example: dx
x.

How do we handle this? We need some basic exponent rules:

√
• Roots are fractional exponents: x1/n = n x

• Fractions are negative exponents: x−n = 1

xn
.

So this is really d 1/2

dx
x = (1/2)x1/2−1 = 1/2x−1/2 = 1
√
2 x
.
√
d
x2 .
3
Example: dx

Using those same exponent rules, this is (x2 )1/3 = x2/3 . So the derivative will be 23 x2/3−1 =
2 −1/3
3
x = 3√
3 x.
2

Continuous vs Differentiable
Recall that a function is continous at a if limx→a f (x) = f (a), and differentiable if f ′ (a) is a
number. Last class, we gave examples of continuous functions that were not differentiable:

12
 √
|x| and 3 x. It turns out that you can’t have the opposite: if f is not continuous, then it is
not differentiable. Equivalently, if f is differentiable, then it is continuous.

You can prove this in detail with limits, but I’ll just give an informal explanation:

If a function has a jump discontinuity, the tangent line from one or both sides will always
tend towards vertical.

13