Differentiability and Continuity
A function is differentiable at a point when it has a derivative there, which means its graph has a single, non-vertical tangent line. Most functions you meet are differentiable almost everywhere, but not quite everywhere. This page shows exactly where derivatives break down, and how differentiability connects to continuity.
Key ideas
Section titled “Key ideas”What differentiable means
Section titled “What differentiable means”is differentiable at if the limit
exists as a finite number. Like any limit, it exists only if the left-hand limit () and the right-hand limit () are equal. These one-sided limits are the slopes from the left and from the right.
A function is differentiable on an interval if it is differentiable at every point of the interval.
Differentiable implies continuous (but not the other way)
Section titled “Differentiable implies continuous (but not the other way)”- If is differentiable at , then is continuous at .
- The reverse is false: a function can be continuous at a point and still not be differentiable there. The classic example is at .
Flip the first statement around and you get a useful test: if is not continuous at , then is not differentiable at .
Four ways a derivative can fail to exist
Section titled “Four ways a derivative can fail to exist”| Type | What the graph looks like | What goes wrong |
|---|---|---|
| Corner | two straight-ish pieces meet at an angle | the slopes from the left and right are different numbers |
| Cusp | a sharp point | the slopes go to on one side and on the other |
| Vertical tangent | the graph is momentarily vertical | the slopes go to (or ) from both sides |
| Discontinuity | a hole, jump, or asymptote | not continuous, so not differentiable |
On the AP exam you usually don’t need the name. You need the reason: ” is not differentiable at because the slopes from the left and right are not equal” or “because is not continuous at ”.
Checking a piecewise function
Section titled “Checking a piecewise function”For a piecewise function made of “nice” pieces (like polynomials), check the boundary point in two steps:
- Continuity: the two pieces must give the same value at (and that must be ).
- Matching slopes: the derivatives of the two pieces must give the same value at .
If both hold, is differentiable at . If step 1 fails, stop: is not differentiable, no matter what the slopes do.
For step 2 you need derivatives of the pieces. From the definition, , , and . The power rule makes these quick.
Worked examples
Section titled “Worked examples”Example 1: A corner
Section titled “Example 1: A corner”Show that is continuous but not differentiable at .
Solution. and , so is continuous at .
The difference quotient at is .
- From the right (): , so the limit is .
- From the left (): , so the limit is .
The one-sided limits are different, so does not exist. The graph has a corner at .
Example 2: A vertical tangent and a cusp
Section titled “Example 2: A vertical tangent and a cusp”Use difference quotients at to explain the shapes of and in the figure.
Solution. For :
Since is positive for , this goes to from both sides. The slopes blow up the same way on both sides, so there is a vertical tangent at .
For :
This goes to as and to as , so there is a cusp at . Neither function is differentiable at .
Example 3: A piecewise function that is differentiable
Section titled “Example 3: A piecewise function that is differentiable”Is differentiable at ?
Solution.
- Continuity: the left piece gives and the right piece gives . They match, and , so is continuous at .
- Slopes: the left piece has derivative , which is at . The right piece has derivative . They match.
So is differentiable at , and . (The line is the tangent line to at , so the pieces join smoothly.)
Example 4: Solving for constants
Section titled “Example 4: Solving for constants”Find and so that is differentiable at .
Solution. Differentiable means continuous and matching slopes, so you get two equations.
Continuity at : .
Slopes at : the derivatives are and , so .
Substitute into the first equation:
Check: both pieces give at , and both slopes are .
Common mistakes
Section titled “Common mistakes”Thinking continuous means differentiable. It doesn’t. is continuous everywhere but has no derivative at . The true statement only goes one way: differentiable implies continuous.
Checking only that the slopes match. If the pieces don’t meet, the function isn’t continuous, so it can’t be differentiable, even if the slopes of the pieces happen to agree. Always check continuity first.
Calling a vertical tangent “differentiable”. The graph does have a tangent line, but it’s vertical, so its slope (the derivative) is undefined.
Trusting the calculator at a sharp point. A graphing calculator’s numerical derivative uses a symmetric difference quotient, so it reports for the “derivative” of at . The real derivative doesn’t exist. On calculator-active AP questions, think about whether the function is differentiable before trusting the number.
Giving a name instead of a reason. “It’s a corner” earns less than “the left-hand and right-hand slopes are and , which are not equal, so does not exist.”
Practice
Section titled “Practice”1. (Warm-up) True or false? Explain.
- (a) If is differentiable at , then is continuous at .
- (b) If is continuous at , then is differentiable at .
- (c) If is not continuous at , then is not differentiable at .
Solution
(a) True. Differentiability implies continuity.
(b) False. For example, is continuous at but has a corner there.
(c) True. This is the flipped (contrapositive) form of (a).
2. (Warm-up) Where is not differentiable? What does the graph look like there?
Solution
At . The graph is a V shape with a corner at : the slope is to the left and to the right.
3. (Warm-up) Explain why is not differentiable at .
Solution
is undefined (division by zero), so is not continuous at . A function that isn’t continuous at a point can’t be differentiable there. (The graph is the line with a hole at .)
4. (Core) Is differentiable at ? Justify.
Solution
Continuity: the left piece gives and the right piece gives , and . Continuous.
Slopes: the left piece has derivative , which is at . The right piece has derivative . Since , is not differentiable at (there’s a corner).
5. (Core) Is differentiable at ? If so, find .
Solution
Continuity: and , and . Continuous.
Slopes: gives at , and the line has slope . They match.
So is differentiable at and .
6. (Core) Let . A student says: “The slopes are and , so is differentiable at .” Is the student right?
Solution
No. Check continuity first: the left piece gives but the right piece gives . The graph jumps, so is not continuous at , and therefore not differentiable there. Matching slopes aren’t enough.
7. (Core) Use the definition of the derivative to show that is differentiable at , and find .
Solution
The limit exists (it is from both sides), so is differentiable at and .
8. (Challenge) Find and so that is differentiable everywhere.
Solution
Each piece is a polynomial, so only needs checking.
Continuity: , so .
Slopes: the derivatives are and , so .
Subtract the first equation from the second: . Then .
Check: and ; slopes and .
9. (Challenge) Where is not differentiable? Justify using one-sided slopes at one of those points.
Solution
changes sign at , so those are the only candidates. Away from them, is either or , which are polynomials (so differentiable).
At : to the right, with slope . To the left (between and ), with slope . The one-sided slopes and are not equal, so is not differentiable at . By symmetry, the same happens at : the slope from the left is and from the right is .
So is not differentiable at and . It is continuous at both, so these are corners.