Skip to content
Tutorials Center
My learning

Topics

Lesson 29 of 37 · The Riemann Integral

The Fundamental Theorem of Calculus

Prove both parts of the Fundamental Theorem of Calculus for Riemann integrals, see why F(x) = integral of f is continuous, and check the theorem numerically.

Differentiation and integration look like unrelated limits: one of slopes of secants, the other of sums of areas. The Fundamental Theorem of Calculus (FTC) shows they are inverse operations. In analysis it comes in two parts, each with precise hypotheses. Knowing them separates correct uses from the common mistake of "integrating anything that has an antiderivative".

Setup: the accumulation function

Let f be Riemann integrable on [a, b]. Define

Text
F(x) = ∫ₐˣ f(t) dt     for x ∈ [a, b].

Lemma. F is continuous on [a, b]. In fact it is Lipschitz.

Proof. Since f is integrable it is bounded, say |f| ≤ B. For x < y, F(y) - F(x) = ∫ₓʸ f, so |F(y) - F(x)| ≤ B(y - x). ∎

So an integral smooths a function: even if f has jumps, F has none.

FTC, Part 1

Theorem. If f is integrable on [a, b] and continuous at a point c ∈ [a, b], then F is differentiable at c and F'(c) = f(c).

Proof. For h > 0 (the case h < 0 is the same),

Text
F(c + h) - F(c)   1   c+h
--------------- = - ∫   f(t) dt.
       h          h   c

Subtract f(c) = (1/h) ∫ᶜ^{c+h} f(c) dt:

Text
| (F(c+h) - F(c))/h - f(c) | = | (1/h) ∫ᶜ^{c+h} ( f(t) - f(c) ) dt | ≤ sup_{t ∈ [c, c+h]} |f(t) - f(c)|.

Given ε > 0, continuity at c gives δ with |f(t) - f(c)| < ε for |t - c| < δ; so the right side is at most ε when 0 < h < δ. ∎

If f is continuous on all of [a, b], then F' = f everywhere: F is an antiderivative of f.

FTC, Part 2

Theorem. If f is integrable on [a, b] and G is differentiable on [a, b] with G' = f, then

Text
∫ₐᵇ f(x) dx = G(b) - G(a).

Proof. Let P = {x₀, ..., xₙ} be any partition. By the Mean Value Theorem, for each i there is tᵢ ∈ (x_{i-1}, xᵢ) with G(xᵢ) - G(x_{i-1}) = f(tᵢ) Δxᵢ. Summing (the left side telescopes):

Text
G(b) - G(a) = Σ f(tᵢ) Δxᵢ.

Since mᵢ ≤ f(tᵢ) ≤ Mᵢ, we get L(f, P) ≤ G(b) - G(a) ≤ U(f, P) for every partition. The only number squeezed between all lower and all upper sums is the integral, hence G(b) - G(a) = ∫ f. ∎

Both hypotheses matter: f must be integrable, and G' = f must hold at every point. A function can have a derivative that is not Riemann integrable (it can be unbounded), in which case the formula has no meaning.

Python Numerical check: F'(x) = f(x) and F(b) - F(a)

Press Run (or Ctrl + Enter) to see the result here.

Program input · what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

What happens at a jump

If f is not continuous at c, Part 1 can fail. Take f(x) = 0 for x < 0, f(x) = 1 for x ≥ 0, and F(x) = ∫₋₁ˣ f. Then F(x) = 0 for x ≤ 0 and F(x) = x for x ≥ 0. The function F is continuous, but at 0 the left derivative is 0 and the right derivative is 1, so F is not differentiable there. (Indeed F(x) = max(x, 0).)

Python The integral of a step function has a corner

Press Run (or Ctrl + Enter) to see the result here.

Program input · what you'd type while the program runs

Type the answers before you press Run, one per line. Each time the program reads a line (with input()), it gets the next line from this box.

Here the formula for F comes from direct area reasoning, not from a numerical integration routine.

Consequences you will use

  • Integration by parts and substitution are proved by applying Part 2 to the product rule and chain rule.
  • Every continuous function on an interval has an antiderivative, namely F. That antiderivative may not be an elementary formula (for example ∫ cos(t²) dt).
  • Mean value theorem for integrals. If f is continuous on [a, b], there is c ∈ [a, b] with ∫ₐᵇ f = f(c)(b - a). It follows from the Extreme and Intermediate Value Theorems, or from MVT applied to F.

Check your understanding

Quick check · 4 questions

of 4 answered You scored / 4 · perfect!

  1. 1. If f is integrable, the function F(x) = integral of f from a to x is always:

    |F(y) - F(x)| is at most B|y - x| where B bounds |f|.
  2. 2. FTC Part 1 gives F'(c) = f(c) provided:

    Continuity at c lets the average of f over [c, c+h] approach f(c).
  3. 3. Which step of the proof of FTC Part 2 uses the Mean Value Theorem?

    MVT applied to G on each subinterval produces the points t_i.
  4. 4. For the step function f = 0 on x below 0 and 1 for x at least 0, F(x) is differentiable at 0?

    F(x) = max(x, 0) has a corner at 0.

Summary

  • F(x) = ∫ₐˣ f is continuous whenever f is integrable.
  • FTC 1: if f is continuous at c, then F'(c) = f(c).
  • FTC 2: if G' = f with f integrable, then ∫ₐᵇ f = G(b) - G(a); the proof squeezes G(b) - G(a) between all lower and upper sums.
  • At a jump of f, F has a corner, so continuity of f is needed in Part 1.

Next: Improper Integrals and Lebesgue's Criterion