MATH 203 Calculus I • Concordia University, Montreal

Corrected exercises: monotonic functions and the First Derivative Test (MATH 203)

This is the corrected exercise set for section 4.3 of Thomas in MATH 203, Differential and Integral Calculus I, at Concordia University: monotonic functions and the First Derivative Test for Local Extrema. The chapter asks two questions about a function, where it increases or decreases and where it has a local maximum or minimum, and answers both with the SIGN of f′f'. Thomas proves the tool from the Mean Value Theorem; MATH 203 skips that proof and uses the result, with its hypotheses, which is what every solution below does. A scientific calculator is allowed, but the answers stay exact whenever they are naturally exact.

The thread running through the set: a sign is only readable on f′f' written as ONE fraction, FULLY factored. The test itself takes three lines; the marks are lost before it, in the algebra that produces the factored form. So each solution names that algebraic gesture where it happens: the common factor taken out first, the minus sign of the quotient rule applied to a whole bracket, the lowest power factored out of x5/3x^{5/3} and x−1/3x^{-1/3}, the identity sin⁡2x=2sin⁡xcos⁡x\sin 2x = 2\sin x\cos x followed by a factorization, a common denominator over xx or over (1+x)2(1 + x)^2.

The traps named in the solutions: alternating signs across a squared factor, forgetting a factor taken out before solving a quadratic, the minus sign of the quotient rule applied to the first term only, a critical point lost by multiplying or dividing by x1/3x^{1/3} or by cos⁡x\cos x, a sign change at a number outside the domain read as an extremum, the endpoints of a closed interval left out of the list of local extrema, an inequality multiplied through by a denominator of unknown sign, and a proof by monotonicity that forgets its starting value.

10 corrected exercises • 100 points • 150 minutes

Revision sheet for this chapter → Every MATH 203 chapter →

Self-checking set Type your answers under each question: the page tells you right or wrong before you open the solution. With an account, every answer right on the first try earns points.

Course recap

  • • Corollary (Thomas 4.3): ff continuous on [a,b][a, b], differentiable on (a,b)(a, b). If f′>0f' > 0 on (a,b)(a, b), ff is increasing on [a,b][a, b]; if f′<0f' < 0 on (a,b)(a, b), decreasing on [a,b][a, b].
  • • Critical point: cc in the domain of ff with f′(c)=0f'(c) = 0 or f′(c)f'(c) undefined.
  • • First Derivative Test at a critical point cc where ff is continuous: f′f' from −- to ++, local minimum; from ++ to −-, local maximum; no sign change, no local extremum.
  • • Local extremum at an endpoint cc of the domain: compare f(c)f(c) with f(x)f(x) for the points xx of the domain near cc; the sign of f′f' next to cc decides.
  • • Sign chart: f′f' as one fraction, numerator and denominator factored; one column per zero AND per point outside the domain; a squared factor or an always-positive factor never changes the sign.
  • • Fractional powers: factor out the smallest exponent, x5/3−4x−1/3=x−1/3(x2−4)x^{5/3} - 4x^{-1/3} = x^{-1/3}(x^2 - 4).

Part A: the basics (/50)

Exercise 1: Polynomials: factor f' completely, then read the sign factor by factor

Thomas states the tool as a corollary of the Mean Value Theorem, which MATH 203 uses without proving it: if ff is continuous on [a,b][a, b] and differentiable on (a,b)(a, b), and f′(x)>0f'(x) > 0 at each point of (a,b)(a, b), then ff is increasing on [a,b][a, b]; if f′(x)<0f'(x) < 0 there, ff is decreasing on [a,b][a, b].

First Derivative Test for Local Extrema: at a critical point cc where ff is continuous, if f′f' changes from negative to positive, ff has a local minimum at cc; from positive to negative, a local maximum; if f′f' keeps its sign, no local extremum. Everything therefore rests on the SIGN of f′f', and a sign is read on a product, one factor at a time. Expanded, f′f' tells you nothing.

  • a) Let f(x)=x3−3x2−24x+4f(x) = x^3 - 3x^2 - 24x + 4. Find the intervals on which ff is increasing and decreasing, and the local maximum and minimum values of ff.
  • b) Let g(x)=3x4+4x3+6x2−5g(x) = 3x^4 + 4x^3 + 6x^2 - 5. Factor g′(x)g'(x) as far as possible over the real numbers, and find the local extreme values of gg.
  • c) Let h(x)=x5−5x4+5x3h(x) = x^5 - 5x^4 + 5x^3. Find the critical points of hh and classify each one with the First Derivative Test.
  • d) For c), a student factors h′h' correctly, then fills the sign row from the right, alternating: ++ after 33, −- on (1,3)(1, 3), ++ on (0,1)(0, 1), −- before 00, and announces a local minimum at 00. Correct the chart.

Type your answers, the page tells you right or wrong 0/9

a)
ff decreases on ,
b)
c)
d)
Show the solution

Answers

  • a) f′(x)=3(x−4)(x+2)f'(x) = 3(x - 4)(x + 2): increasing on (−∞,−2](-\infty, -2] and [4,∞)[4, \infty), decreasing on [−2,4][-2, 4]; local max f(−2)=32f(-2) = 32, local min f(4)=−76f(4) = -76
  • b) g′(x)=12x(x2+x+1)g'(x) = 12x(x^2 + x + 1), the quadratic has no real root; local min g(0)=−5g(0) = -5, no local max
  • c) h′(x)=5x2(x−1)(x−3)h'(x) = 5x^2(x - 1)(x - 3): critical points 00, 11, 33; local max h(1)=1h(1) = 1, local min h(3)=−27h(3) = -27, nothing at 00
  • d) 5x2>05x^2 > 0 on both sides of 00: h′>0h' > 0 on (−∞,0)(-\infty, 0) AND on (0,1)(0, 1), so no extremum at 00.

a) ff is a polynomial, so f′f' exists everywhere and the critical points are the zeros of f′f'. The algebraic gesture: take out the common factor FIRST, f′(x)=3x2−6x−24=3(x2−2x−8)f'(x) = 3x^2 - 6x - 24 = 3(x^2 - 2x - 8), then factor the trinomial by looking for two numbers with product −8-8 and sum −2-2: f′(x)=3(x−4)(x+2)f'(x) = 3(x - 4)(x + 2). The factor 33 is positive and never matters. On (−∞,−2)(-\infty, -2) both brackets are negative, f′>0f' > 0; on (−2,4)(-2, 4) only x−4x - 4 is negative, f′<0f' < 0; on (4,∞)(4, \infty) both are positive, f′>0f' > 0. By the corollary, ff is increasing on (−∞,−2](-\infty, -2], decreasing on [−2,4][-2, 4] and increasing on [4,∞)[4, \infty) (the endpoints may be included since ff is continuous). First Derivative Test: ++ to −- at −2-2, local maximum f(−2)=−8−12+48+4=32f(-2) = -8 - 12 + 48 + 4 = 32; −- to ++ at 44, local minimum f(4)=64−48−96+4=−76f(4) = 64 - 48 - 96 + 4 = -76.

b) g′(x)=12x3+12x2+12x=12x(x2+x+1)g'(x) = 12x^3 + 12x^2 + 12x = 12x(x^2 + x + 1). Now the trinomial: its discriminant is 12−4(1)(1)=−3<01^2 - 4(1)(1) = -3 < 0, so it has no real root and it keeps the sign of its value at 00, which is 11: x2+x+1>0x^2 + x + 1 > 0 for every xx (completing the square, (x+12)2+34\left(x + \frac{1}{2}\right)^2 + \frac{3}{4}). This factor is therefore dropped from the sign, never from the derivative: the sign of g′g' is the sign of 12x12x. The only critical point is 00, gg is decreasing on (−∞,0](-\infty, 0] and increasing on [0,∞)[0, \infty), and the First Derivative Test gives a local minimum g(0)=−5g(0) = -5, and no local maximum at all. The classic loss: the student sees that x2+x+1=0x^2 + x + 1 = 0 has no solution, writes “g′g' has no zero, no critical point”, and forgets the factor 12x12x that was taken out first.

c) h′(x)=5x4−20x3+15x2=5x2(x2−4x+3)=5x2(x−1)(x−3)h'(x) = 5x^4 - 20x^3 + 15x^2 = 5x^2(x^2 - 4x + 3) = 5x^2(x - 1)(x - 3), with the lowest power x2x^2 taken out as a common factor. Critical points: 00, 11 and 33. The sign chart of the solution figure takes each factor on each interval. 5x25x^2 is positive except at 00; x−1x - 1 and x−3x - 3 change sign at their zeros. So h′>0h' > 0 on (−∞,0)(-\infty, 0) and on (0,1)(0, 1), h′<0h' < 0 on (1,3)(1, 3), h′>0h' > 0 on (3,∞)(3, \infty). Hence hh is increasing on (−∞,1](-\infty, 1] (the two intervals share the point 00), decreasing on [1,3][1, 3], increasing on [3,∞)[3, \infty). At 11: ++ to −-, local maximum h(1)=1−5+5=1h(1) = 1 - 5 + 5 = 1. At 33: −- to ++, local minimum h(3)=243−405+135=−27h(3) = 243 - 405 + 135 = -27. At 00: ++ on both sides, no local extremum, although h′(0)=0h'(0) = 0.

d) Signs alternate only across a zero of ODD multiplicity. The factor x2x^2 is a square: it is positive on both sides of 00 and cannot flip the sign of h′h'. The row written by rhythm, −∣+∣−∣+- \mid + \mid - \mid +, is wrong on (−∞,0)(-\infty, 0): a test value confirms it, h′(−1)=5(1)(−2)(−4)=40>0h'(-1) = 5(1)(-2)(-4) = 40 > 0. The correct row is +∣+∣−∣++ \mid + \mid - \mid +, and the point 00 is a critical point without extremum, a horizontal tangent on a rising curve. The student's “local minimum at 00” is a point that does not exist, and a marker takes the whole classification away. The safe method: one row per factor, as in the figure, or one test value in EVERY interval, never a pattern written in advance.

x0135x²+0+++x − 1−−0++x − 3−−−0+h'(x)+0+0−0+h(x)↗0↗1↘−27↗

Tick the exercises you have done or want to review: a free account, no password, keeps your ticks from one visit to the next and tells you which chapter to tackle next. Create your space, an email is enough.

Exercise 2: Quotients: the minus sign of the quotient rule, and the holes of the domain

For a quotient, (uv)′=u′v−uv′v2\left(\frac{u}{v}\right)' = \frac{u'v - uv'}{v^2}, and the minus sign applies to the WHOLE product uv′uv': write it with brackets before expanding. The denominator v2v^2 is positive wherever ff is defined, so the sign of f′f' is the sign of the numerator, once it is factored.

A number where ff is not defined is not a critical point, yet it must be a column of the sign chart: f′f' can change sign there with no extremum at all. The figure shows the graph of f(x)=x2+5xx−4f(x) = \frac{x^2 + 5x}{x - 4} and its vertical asymptote x=4x = 4.

-10-8-6-4-224681012141618-16-8816243240(−2, 1)(10, 25)y = f(x)x = 4x
  • a) Show that f′(x)=(x−10)(x+2)(x−4)2f'(x) = \frac{(x - 10)(x + 2)}{(x - 4)^2}. Find the intervals on which ff is increasing and decreasing, and its local extreme values.
  • b) On the figure, the local maximum value is SMALLER than the local minimum value. Explain why this is not a contradiction.
  • c) Let g(x)=x(x−2)2g(x) = \frac{x}{(x - 2)^2}. Show that g′(x)=−x+2(x−2)3g'(x) = -\frac{x + 2}{(x - 2)^3}, find the intervals of increase and decrease, and the local extrema of gg.
  • d) A student starts a) with f′(x)=(2x+5)(x−4)−x2+5x(x−4)2=x2+2x−20(x−4)2f'(x) = \frac{(2x + 5)(x - 4) - x^2 + 5x}{(x - 4)^2} = \frac{x^2 + 2x - 20}{(x - 4)^2}. Find the error and say what it does to the answer.

Type your answers, the page tells you right or wrong 0/9

a)
ff increases on the right of ,
b)
c)
d)
Show the solution

Answers

  • a) Increasing on (−∞,−2](-\infty, -2] and [10,∞)[10, \infty), decreasing on [−2,4)[-2, 4) and (4,10](4, 10]; local max f(−2)=1f(-2) = 1, local min f(10)=25f(10) = 25
  • b) Local means near the point: the two extrema lie on the two branches separated by the asymptote x=4x = 4.
  • c) Decreasing on (−∞,−2](-\infty, -2] and (2,∞)(2, \infty), increasing on [−2,2)[-2, 2); local min g(−2)=−18g(-2) = -\frac{1}{8}, no local max (22 is not in the domain)
  • d) The minus sign was not applied to 5x5x: the numerator is x2−8x−20x^2 - 8x - 20, not x2+2x−20x^2 + 2x - 20, and the critical points −1±21-1 \pm \sqrt{21} are wrong.

a) With u=x2+5xu = x^2 + 5x and v=x−4v = x - 4: f′(x)=(2x+5)(x−4)−(x2+5x)(1)(x−4)2f'(x) = \frac{(2x + 5)(x - 4) - (x^2 + 5x)(1)}{(x - 4)^2}. Brackets first, then expand: (2x+5)(x−4)=2x2−3x−20(2x + 5)(x - 4) = 2x^2 - 3x - 20, and −(x2+5x)=−x2−5x-(x^2 + 5x) = -x^2 - 5x. The numerator is x2−8x−20=(x−10)(x+2)x^2 - 8x - 20 = (x - 10)(x + 2). The domain is x≠4x \ne 4, and (x−4)2>0(x - 4)^2 > 0 on it, so the sign of f′f' is that of (x−10)(x+2)(x - 10)(x + 2): positive for x<−2x < -2, negative on (−2,4)(-2, 4) and on (4,10)(4, 10), positive for x>10x > 10. The chart has THREE columns, −2-2, 44 and 1010, the middle one being a double bar. So ff is increasing on (−∞,−2](-\infty, -2], decreasing on [−2,4)[-2, 4) and on (4,10](4, 10] (separately: never “on [−2,10][-2, 10]”, which contains 44), increasing on [10,∞)[10, \infty). Local maximum f(−2)=4−10−6=1f(-2) = \frac{4 - 10}{-6} = 1, local minimum f(10)=100+506=25f(10) = \frac{100 + 50}{6} = 25.

b) A local maximum is only larger than the values NEAR it, and a local minimum only smaller than the values near it. Here the maximum at x=−2x = -2 lives on the left branch and the minimum at x=10x = 10 on the right branch; between them ff is not even defined at 44, where the left branch falls to −∞-\infty and the right branch comes down from +∞+\infty. Nothing forces the two heights to be in any order. On an interval where ff is continuous, two consecutive extrema of opposite types are ordered; across an asymptote, no rule applies.

c) Quotient rule with u=xu = x, v=(x−2)2v = (x - 2)^2, v′=2(x−2)v' = 2(x - 2) (chain rule): g′(x)=(x−2)2−x⋅2(x−2)(x−2)4g'(x) = \frac{(x - 2)^2 - x \cdot 2(x - 2)}{(x - 2)^4}. The gesture: do NOT expand. Take out the common factor (x−2)(x - 2) of the numerator and cancel it once: g′(x)=(x−2)[(x−2)−2x](x−2)4=−x−2(x−2)3=−x+2(x−2)3g'(x) = \frac{(x - 2)\left[(x - 2) - 2x\right]}{(x - 2)^4} = \frac{-x - 2}{(x - 2)^3} = -\frac{x + 2}{(x - 2)^3}. This time the denominator is a CUBE, so it changes sign at 22. For x<−2x < -2: −(x+2)>0-(x + 2) > 0 over a negative cube, g′<0g' < 0. On (−2,2)(-2, 2): −(x+2)<0-(x + 2) < 0 over a negative cube, g′>0g' > 0. For x>2x > 2: negative over positive, g′<0g' < 0. So gg is decreasing on (−∞,−2](-\infty, -2], increasing on [−2,2)[-2, 2), decreasing on (2,∞)(2, \infty). Local minimum g(−2)=−216=−18g(-2) = \frac{-2}{16} = -\frac{1}{8}. At 22, g′g' goes from ++ to −-, but 22 is not in the domain: no maximum, the graph climbs to +∞+\infty on both sides.

d) The student wrote −x2+5x- x^2 + 5x instead of −(x2+5x)=−x2−5x-(x^2 + 5x) = -x^2 - 5x: the minus sign was applied to the first term of uv′uv' only. The numerator becomes 2x2−3x−20−x2+5x=x2+2x−202x^2 - 3x - 20 - x^2 + 5x = x^2 + 2x - 20, whose zeros are −1±21-1 \pm \sqrt{21}, about −5.58-5.58 and 3.583.58: two false critical points, a false chart, and every conclusion of a) lost, although the quotient rule itself was known. A check catches it, but not at x=0x = 0: there the faulty term 5x5x vanishes and both formulas give −2016=−54-\frac{20}{16} = -\frac{5}{4}. At x=10x = 10 the true value is 00 and the student's is 10036=259\frac{100}{36} = \frac{25}{9}. Test the simplified derivative against the unsimplified one at two easy points, one of them away from 00.

Exercise 3: Fractional powers and roots: factor out the LOWEST power

When f′f' is a sum of powers such as x5/3x^{5/3} and x−1/3x^{-1/3}, or contains a square root, it becomes readable only as ONE fraction. The gesture: factor out the power with the smallest exponent (the most negative one), even if it is negative, then send it to the denominator. Nothing is ever divided away: a number where f′f' does not exist but ff does is a critical point.

Thomas defines a local extremum at an ENDPOINT of the domain too: ff has a local minimum at an endpoint cc if f(x)≥f(c)f(x) \ge f(c) for the points xx of the domain near cc. The sign of f′f' next to the endpoint decides it.

  • a) Let f(x)=x2/3(x2−16)f(x) = x^{2/3}(x^2 - 16). Show that f′(x)=8(x−2)(x+2)3x1/3f'(x) = \frac{8(x - 2)(x + 2)}{3x^{1/3}}, and list ALL the critical points of ff.
  • b) Classify each critical point of ff with the First Derivative Test, and give the local extreme values exactly.
  • c) Let g(x)=x6−xg(x) = x\sqrt{6 - x}. Give its domain, show that g′(x)=3(4−x)26−xg'(x) = \frac{3(4 - x)}{2\sqrt{6 - x}}, and find all the local extrema of gg, endpoint included.
  • d) A student solves f′(x)=0f'(x) = 0 in a) by multiplying 83x5/3−323x−1/3=0\frac{8}{3}x^{5/3} - \frac{32}{3}x^{-1/3} = 0 by x1/3x^{1/3}, finds x=±2x = \pm 2 and writes: the critical points are −2-2 and 22. What did this cost?

Type your answers, the page tells you right or wrong 0/8

a)
b)
c)
gg increases on ,
d)
Show the solution

Answers

  • a) Critical points −2-2, 22 (f′=0f' = 0) and 00 (f′f' undefined, f(0)=0f(0) = 0)
  • b) Local min f(±2)=−1243f(\pm 2) = -12\sqrt[3]{4} at x=−2x = -2 and x=2x = 2; local max f(0)=0f(0) = 0, at a cusp
  • c) Domain (−∞,6](-\infty, 6]; increasing on (−∞,4](-\infty, 4], decreasing on [4,6][4, 6]; local max g(4)=42g(4) = 4\sqrt 2, local min g(6)=0g(6) = 0 at the endpoint
  • d) The critical point 00 was lost, and with it the only local maximum of ff.

a) Expand first to have pure powers: f(x)=x8/3−16x2/3f(x) = x^{8/3} - 16x^{2/3}, so f′(x)=83x5/3−323x−1/3f'(x) = \frac{8}{3}x^{5/3} - \frac{32}{3}x^{-1/3}. The exponents are 53\frac{5}{3} and −13-\frac{1}{3}; the smaller one is −13-\frac{1}{3}, and factoring it out SUBTRACTS it from each exponent: x5/3=x−1/3⋅x2x^{5/3} = x^{-1/3} \cdot x^{2}. So f′(x)=83x−1/3(x2−4)=8(x−2)(x+2)3x1/3f'(x) = \frac{8}{3}x^{-1/3}(x^2 - 4) = \frac{8(x - 2)(x + 2)}{3x^{1/3}}. The domain of ff is all real numbers (a cube root accepts negatives). Critical points: the zeros of the numerator, −2-2 and 22, AND 00, where the denominator vanishes, f′(0)f'(0) does not exist, but f(0)=0f(0) = 0 exists.

b) The cube root x1/3x^{1/3} has the sign of xx, and (x−2)(x+2)(x - 2)(x + 2) is positive outside [−2,2][-2, 2], negative inside. For x<−2x < -2: ++ over −-, f′<0f' < 0. On (−2,0)(-2, 0): −- over −-, f′>0f' > 0. On (0,2)(0, 2): −- over ++, f′<0f' < 0. For x>2x > 2: f′>0f' > 0. First Derivative Test: local minima at −2-2 and 22, with f(±2)=41/3(4−16)=−1243f(\pm 2) = 4^{1/3}(4 - 16) = -12\sqrt[3]{4} (about −19.05-19.05), and a local maximum at 00, f(0)=0f(0) = 0. The maximum sits exactly where f′f' does not exist: as x→0−x \to 0^-, f′f' is large and positive, as x→0+x \to 0^+ large and negative, so the graph has a CUSP at the origin, as in the solution figure. The symmetry is a free check: f(−x)=f(x)f(-x) = f(x), the two minima must have the same value.

c) The square root needs 6−x≥06 - x \ge 0: the domain is (−∞,6](-\infty, 6]. Product rule: g′(x)=6−x+x⋅−126−xg'(x) = \sqrt{6 - x} + x \cdot \frac{-1}{2\sqrt{6 - x}}, for x<6x < 6. Common denominator 26−x2\sqrt{6 - x}: the first term becomes 2(6−x)26−x\frac{2(6 - x)}{2\sqrt{6 - x}}, since 6−x⋅6−x=6−x\sqrt{6 - x} \cdot \sqrt{6 - x} = 6 - x. So g′(x)=12−2x−x26−x=3(4−x)26−xg'(x) = \frac{12 - 2x - x}{2\sqrt{6 - x}} = \frac{3(4 - x)}{2\sqrt{6 - x}}. The denominator is positive on (−∞,6)(-\infty, 6): g′>0g' > 0 for x<4x < 4, g′<0g' < 0 on (4,6)(4, 6). Critical points: 44 (g′=0g' = 0) and 66 (g′g' undefined, and an endpoint). gg is increasing on (−∞,4](-\infty, 4] and decreasing on [4,6][4, 6]. Local maximum g(4)=42g(4) = 4\sqrt 2. At the endpoint 66, gg decreases INTO it, so g(x)>g(6)=0g(x) > g(6) = 0 for xx of the domain near 66: local minimum 00 at the endpoint, with a vertical tangent.

d) Multiplying by x1/3x^{1/3} is legitimate only for x≠0x \ne 0, and precisely at 00 the derivative does not exist: the multiplication silently deleted a critical point. It was not a harmless one: 00 is the only local maximum of ff, and its value 00 is part of the expected answer. The correct gesture is the factorization of a), which keeps x1/3x^{1/3} visible in the denominator, so the chart shows the column 00 and its sign change. The same rule holds for any factor: never multiply or divide an equation f′(x)=0f'(x) = 0 by an expression that can be zero or undefined; factor it and look at it.

-5-4-3-2-112345-24-20-16-12-8-448121620min at x = −2min at x = 2cusp: local max (0, 0)x

Exercise 4: Reading f on the graph of f', when f' jumps

The figure shows the graph of the DERIVATIVE f′f' of a function ff that is continuous on [−4,6][-4, 6] and differentiable at every point of (−4,6)(-4, 6) except x=2x = 2. The graph of f′f' is made of two straight pieces through grid points; the open circles at (2,2)(2, 2) and (2,−3)(2, -3) mean that f′(2)f'(2) does not exist.

Only the SIGN of f′f', above or below the axis, speaks about the direction of ff. The height of f′f' says how steep ff is, never how high ff is.

-4-3-2-1123456-4-3-2-112345y = f'(x)(2, 2)(2, −3)x
  • a) List the critical points of ff in (−4,6)(-4, 6), saying for each whether f′=0f' = 0 or f′f' does not exist there.
  • b) On which intervals is ff increasing, decreasing?
  • c) Classify each critical point, and say whether ff has a local extremum at each endpoint x=−4x = -4 and x=6x = 6.
  • d) A student writes: the graph of f′f' is lowest just to the right of 22, at height −3-3, so ff has a local minimum at x=2x = 2. Correct.
  • e) Compare, when the figure allows it: f(0)f(0) and f(2)f(2); f(3)f(3) and f(5)f(5); f(−4)f(-4) and f(−2)f(-2). Can the figure alone decide between f(−4)f(-4) and f(6)f(6)?

Type your answers, the page tells you right or wrong 0/9

a)
b)
ff increases on ,
ff decreases on the interval ,
c)
d)
e)
Show the solution

Answers

  • a) −2-2 and 55 (f′=0f' = 0), and 22 (f′f' does not exist)
  • b) Decreasing on [−4,−2][-4, -2] and [2,5][2, 5], increasing on [−2,2][-2, 2] and [5,6][5, 6]
  • c) Local min at −2-2 and at 55, local max at 22; local max at both endpoints −4-4 and 66
  • d) False: f′f' goes from ++ to −- at 22, so ff has a local MAXIMUM there, at a corner.
  • e) f(0)<f(2)f(0) < f(2), f(3)>f(5)f(3) > f(5), f(−4)>f(−2)f(-4) > f(-2); f(−4)f(-4) against f(6)f(6) is not decided by the signs.

a) A critical point is a number of the domain where f′f' is zero or does not exist. The graph of f′f' meets the axis at x=−2x = -2 (first piece) and at x=5x = 5 (second piece): f′(−2)=f′(5)=0f'(-2) = f'(5) = 0. At x=2x = 2 both pieces end with an open circle: f′(2)f'(2) does not exist, while f(2)f(2) exists because ff is continuous on [−4,6][-4, 6]. So the critical points are −2-2, 22 and 55. Forgetting 22 is the classic loss: the eye looks for crossings of the axis, and at 22 the graph of f′f' does not cross it, it JUMPS over it.

b) Read the sign: below the axis on [−4,−2)[-4, -2), above on (−2,2)(-2, 2), below on (2,5)(2, 5), above on (5,6](5, 6]. By the corollary, applied on each interval where ff is continuous and differentiable inside: ff is decreasing on [−4,−2][-4, -2], increasing on [−2,2][-2, 2], decreasing on [2,5][2, 5] and increasing on [5,6][5, 6]. The point 22 may close both intervals [−2,2][-2, 2] and [2,5][2, 5] because ff is continuous at 22; the corollary only needs differentiability INSIDE each interval.

c) At −2-2: −- to ++, local minimum. At 22: ++ to −-, local maximum (the First Derivative Test only needs ff continuous at 22, not differentiable there). At 55: −- to ++, local minimum. Endpoints: ff is decreasing on [−4,−2][-4, -2], so f(x)<f(−4)f(x) < f(-4) for xx just to the right of −4-4: local maximum at −4-4. ff is increasing on [5,6][5, 6], so f(x)<f(6)f(x) < f(6) just to the left of 66: local maximum at 66. A table of local extrema on a closed interval that stops at the interior points is missing two lines.

d) The student read the graph of f′f' as if it were the graph of ff. The value −3-3 is the slope of ff just after 22: there ff is falling steeply, it does not sit at a bottom. What happens at 22 is read on the SIGN: f′f' is positive just before 22 (values near 22) and negative just after (values near −3-3), so ff rises then falls: a local MAXIMUM at 22. Since the two one-sided slopes are different, 22 and −3-3, the graph of ff has a corner there, a peak with two different tangent directions.

e) ff is increasing on [−2,2][-2, 2] and 0<20 < 2: f(0)<f(2)f(0) < f(2), strictly since f′>0f' > 0 on (0,2)(0, 2). ff is decreasing on [2,5][2, 5] and 3<53 < 5: f(3)>f(5)f(3) > f(5). ff is decreasing on [−4,−2][-4, -2]: f(−4)>f(−2)f(-4) > f(-2). But −4-4 and 66 are separated by increases AND decreases: the signs say where ff goes, not how far. Deciding would require measuring the net change of ff from the areas between the graph of f′f' and the axis, which is the business of MATH 205, not of the First Derivative Test.

Exercise 5: Trigonometric functions on an interval: factor cos x, never divide by it

With trigonometric functions the algebra that costs marks is the identity used too late and the factor divided away. sin⁡2x=2sin⁡xcos⁡x\sin 2x = 2\sin x\cos x and cos⁡2x=1−2sin⁡2x=2cos⁡2x−1\cos 2x = 1 - 2\sin^2 x = 2\cos^2 x - 1 turn f′f' into a product; then EACH factor is set to zero on the interval.

The figure shows f(x)=cos⁡2x+2sin⁡xf(x) = \cos 2x + 2\sin x on [0,2π][0, 2\pi]. The endpoints belong to the interval, so they can carry local extrema in Thomas's sense.

π/6π/25π/63π/22π1−1−3y = cos 2x + 2 sin x
  • a) Show that f′(x)=2cos⁡x (1−2sin⁡x)f'(x) = 2\cos x\,(1 - 2\sin x) and find the critical points of ff in (0,2π)(0, 2\pi).
  • b) Make the sign chart of f′f' on [0,2π][0, 2\pi] and give all the local extreme values of ff, endpoints included, with the points where they occur.
  • c) A student writes −4sin⁡xcos⁡x+2cos⁡x=0-4\sin x\cos x + 2\cos x = 0, divides by 2cos⁡x2\cos x, gets sin⁡x=12\sin x = \frac{1}{2}, and concludes that ff has exactly two critical points. What is lost?
  • d) Let h(x)=2x−sin⁡2xh(x) = 2x - \sin 2x on [0,2π][0, 2\pi]. Show that h′(x)=4sin⁡2xh'(x) = 4\sin^2 x, and deduce that hh is increasing on [0,2π][0, 2\pi] although h′h' vanishes inside the interval.

Type your answers, the page tells you right or wrong 0/8

a)
b)
c)
d)
Show the solution

Answers

  • a) π6\frac{\pi}{6}, π2\frac{\pi}{2}, 5π6\frac{5\pi}{6}, 3π2\frac{3\pi}{2}
  • b) Local max 32\frac{3}{2} at π6\frac{\pi}{6} and 5π6\frac{5\pi}{6}, and 11 at the endpoint 2π2\pi; local min 11 at 00 and at π2\frac{\pi}{2}, −3-3 at 3π2\frac{3\pi}{2}
  • c) The zeros of cos⁡x\cos x, π2\frac{\pi}{2} and 3π2\frac{3\pi}{2}, are lost, among them the lowest point f(3π2)=−3f\left(\frac{3\pi}{2}\right) = -3.
  • d) h′(x)=2−2cos⁡2x=4sin⁡2x≥0h'(x) = 2 - 2\cos 2x = 4\sin^2 x \ge 0, zero only at π\pi inside: hh is increasing on [0,2π][0, 2\pi], no local extremum inside

a) By the chain rule, (cos⁡2x)′=−2sin⁡2x(\cos 2x)' = -2\sin 2x, so f′(x)=−2sin⁡2x+2cos⁡xf'(x) = -2\sin 2x + 2\cos x. The identity is used NOW, on the derivative: sin⁡2x=2sin⁡xcos⁡x\sin 2x = 2\sin x\cos x, so f′(x)=−4sin⁡xcos⁡x+2cos⁡xf'(x) = -4\sin x\cos x + 2\cos x. Both terms contain 2cos⁡x2\cos x: f′(x)=2cos⁡x (1−2sin⁡x)f'(x) = 2\cos x\,(1 - 2\sin x). It is zero when cos⁡x=0\cos x = 0, that is x=π2x = \frac{\pi}{2} or 3π2\frac{3\pi}{2}, or when sin⁡x=12\sin x = \frac{1}{2}, that is x=π6x = \frac{\pi}{6} or 5π6\frac{5\pi}{6}. f′f' exists everywhere: these four numbers are all the critical points in (0,2π)(0, 2\pi).

b) Sign of each factor. cos⁡x>0\cos x > 0 on [0,π2)\left[0, \frac{\pi}{2}\right) and (3π2,2π]\left(\frac{3\pi}{2}, 2\pi\right], <0< 0 between. 1−2sin⁡x>01 - 2\sin x > 0 when sin⁡x<12\sin x < \frac{1}{2}, that is on [0,π6)\left[0, \frac{\pi}{6}\right) and (5π6,2π]\left(\frac{5\pi}{6}, 2\pi\right]. Product: ++ on (0,π6)\left(0, \frac{\pi}{6}\right), −- on (π6,π2)\left(\frac{\pi}{6}, \frac{\pi}{2}\right), ++ on (π2,5π6)\left(\frac{\pi}{2}, \frac{5\pi}{6}\right), −- on (5π6,3π2)\left(\frac{5\pi}{6}, \frac{3\pi}{2}\right), ++ on (3π2,2π)\left(\frac{3\pi}{2}, 2\pi\right). First Derivative Test: local maxima f(π6)=cos⁡π3+2⋅12=32f\left(\frac{\pi}{6}\right) = \cos\frac{\pi}{3} + 2 \cdot \frac{1}{2} = \frac{3}{2} and f(5π6)=cos⁡5π3+1=32f\left(\frac{5\pi}{6}\right) = \cos\frac{5\pi}{3} + 1 = \frac{3}{2}; local minima f(π2)=cos⁡π+2=1f\left(\frac{\pi}{2}\right) = \cos\pi + 2 = 1 and f(3π2)=cos⁡3π−2=−3f\left(\frac{3\pi}{2}\right) = \cos 3\pi - 2 = -3. Endpoints: ff increases right after 00, so local minimum f(0)=1f(0) = 1; ff increases up to 2π2\pi, so local maximum f(2π)=1f(2\pi) = 1. The figure confirms the alternation max, min, max, min between the endpoints.

c) Dividing by 2cos⁡x2\cos x is only allowed where cos⁡x≠0\cos x \ne 0, and precisely the solutions of cos⁡x=0\cos x = 0 are solutions of the equation: π2\frac{\pi}{2} and 3π2\frac{3\pi}{2} disappear. The student then finds two maxima and no minimum inside the interval, which the figure refutes at once, and misses f(3π2)=−3f\left(\frac{3\pi}{2}\right) = -3, the lowest point of the graph. Correct gesture: factor, 2cos⁡x (1−2sin⁡x)=02\cos x\,(1 - 2\sin x) = 0, then “cos⁡x=0\cos x = 0 OR sin⁡x=12\sin x = \frac{1}{2}”, each factor on the interval.

d) h′(x)=2−2cos⁡2x=2(1−cos⁡2x)h'(x) = 2 - 2\cos 2x = 2(1 - \cos 2x), and with cos⁡2x=1−2sin⁡2x\cos 2x = 1 - 2\sin^2 x: h′(x)=2⋅2sin⁡2x=4sin⁡2xh'(x) = 2 \cdot 2\sin^2 x = 4\sin^2 x. It is a square: h′(x)≥0h'(x) \ge 0 everywhere, and in (0,2π)(0, 2\pi) it is zero only at x=πx = \pi. So h′>0h' > 0 on (0,π)(0, \pi) and on (π,2π)(\pi, 2\pi): by the corollary hh is increasing on [0,π][0, \pi] and on [π,2π][\pi, 2\pi], hence on [0,2π][0, 2\pi]. The critical point π\pi is not an extremum, since the sign does not change; the only local extrema are at the endpoints, a local minimum h(0)=0h(0) = 0 and a local maximum h(2π)=4πh(2\pi) = 4\pi. A derivative that vanishes at isolated points does not stop a function from increasing.

Part B: problems and reasoning (/50)

Exercise 6: How many solutions? Monotonic pieces, then the Intermediate Value Theorem

On an interval where ff is continuous and strictly monotonic, ff takes each value AT MOST once: two different points give two different values. The Intermediate Value Theorem then gives AT LEAST one solution of f(x)=kf(x) = k when kk lies between the values of ff at the ends of the piece. Together: exactly one solution per piece whose range contains kk.

The figure shows p(x)=2x3+3x2−36xp(x) = 2x^3 + 3x^2 - 36x and a horizontal line y=ky = k. The number of solutions of p(x)=kp(x) = k is the number of times the line meets the curve, and the point is to count them without trusting the drawing.

-6-5-4-3-2-112345-60-40-2020406080100y = ky = p(x)x
  • a) Find the intervals on which pp is increasing and decreasing, and its local extreme values.
  • b) Give lim⁡x→−∞p(x)\lim_{x \to -\infty} p(x) and lim⁡x→∞p(x)\lim_{x \to \infty} p(x), then the set of values taken by pp on each of the three intervals of a).
  • c) According to the value of the constant kk, give the number of real solutions of p(x)=kp(x) = k.
  • d) Solve p(x)=0p(x) = 0 exactly, and check that the three roots fall one in each interval of a), as c) predicts.
  • e) For which values of kk does p(x)=kp(x) = k have exactly one solution in [−3,2][-3, 2]? Justify.

Type your answers, the page tells you right or wrong 0/10

a)
pp decreases on ,
b)
Values taken on [−3,2][-3, 2] ,
c)
Values of kk with three solutions ,
d)
e)
Values of kk ,
Show the solution

Answers

  • a) p′(x)=6(x+3)(x−2)p'(x) = 6(x + 3)(x - 2): increasing on (−∞,−3](-\infty, -3] and [2,∞)[2, \infty), decreasing on [−3,2][-3, 2]; local max p(−3)=81p(-3) = 81, local min p(2)=−44p(2) = -44
  • b) −∞-\infty and +∞+\infty; values (−∞,81](-\infty, 81], then [−44,81][-44, 81], then [−44,∞)[-44, \infty)
  • c) k>81k > 81 or k<−44k < -44: one solution; k=81k = 81 or k=−44k = -44: two; −44<k<81-44 < k < 81: three
  • d) x=0x = 0 and x=−3±3334x = \frac{-3 \pm 3\sqrt{33}}{4}, about −5.06-5.06 and 3.563.56
  • e) Exactly when −44≤k≤81-44 \le k \le 81

a) p′(x)=6x2+6x−36=6(x2+x−6)=6(x+3)(x−2)p'(x) = 6x^2 + 6x - 36 = 6(x^2 + x - 6) = 6(x + 3)(x - 2), common factor 66 taken out first. Positive outside [−3,2][-3, 2], negative inside. So pp is increasing on (−∞,−3](-\infty, -3], decreasing on [−3,2][-3, 2], increasing on [2,∞)[2, \infty). Local maximum p(−3)=−54+27+108=81p(-3) = -54 + 27 + 108 = 81, local minimum p(2)=16+12−72=−44p(2) = 16 + 12 - 72 = -44. The substitution is where the marks go: 2(−3)3=−542(-3)^3 = -54, and −36(−3)=+108-36(-3) = +108, two signs to keep.

b) The leading term 2x32x^3 dominates (chapter 5): p(x)→−∞p(x) \to -\infty as x→−∞x \to -\infty and p(x)→+∞p(x) \to +\infty as x→∞x \to \infty. On (−∞,−3](-\infty, -3], pp is continuous and increasing from −∞-\infty up to 8181: by the Intermediate Value Theorem it takes every value of (−∞,81](-\infty, 81]. On [−3,2][-3, 2] it decreases from 8181 to −44-44: it takes every value of [−44,81][-44, 81]. On [2,∞)[2, \infty) it increases from −44-44 to +∞+\infty: values [−44,∞)[-44, \infty). On each piece, strict monotonicity says that each of these values is taken exactly once.

c) Count the pieces whose set of values contains kk. If k>81k > 81: only the third piece, one solution. If k=81k = 81: the first piece (at x=−3x = -3), the second (the same point x=−3x = -3!) and the third. Careful, x=−3x = -3 belongs to the first two pieces, so it counts once: two solutions, x=−3x = -3 and one in (2,∞)(2, \infty). If −44<k<81-44 < k < 81: one solution in each piece, and none at a shared endpoint, so three. If k=−44k = -44: x=2x = 2 (shared by the last two pieces) and one in (−∞,−3)(-\infty, -3), two solutions. If k<−44k < -44: first piece only, one solution. The shared endpoints are the trap: counting a solution twice gives “three” at k=81k = 81.

d) p(x)=x(2x2+3x−36)p(x) = x(2x^2 + 3x - 36), and the quadratic formula gives x=−3±9+2884=−3±2974=−3±3334x = \frac{-3 \pm \sqrt{9 + 288}}{4} = \frac{-3 \pm \sqrt{297}}{4} = \frac{-3 \pm 3\sqrt{33}}{4}, since 297=9⋅33297 = 9 \cdot 33. With the calculator, 33≈5.745\sqrt{33} \approx 5.745: the roots are about −5.06-5.06, 00 and 3.563.56. The value k=0k = 0 lies in (−44,81)(-44, 81), so c) announces three solutions, one per piece: indeed −5.06<−3-5.06 < -3, −3<0<2-3 < 0 < 2 and 3.56>23.56 > 2.

e) On [−3,2][-3, 2], pp is continuous and strictly decreasing from p(−3)=81p(-3) = 81 to p(2)=−44p(2) = -44. If −44≤k≤81-44 \le k \le 81, the Intermediate Value Theorem gives at least one solution in [−3,2][-3, 2], and strict monotonicity at most one: exactly one. If k>81k > 81 or k<−44k < -44, kk is not between the values taken on [−3,2][-3, 2], so there is none. Answer: exactly when −44≤k≤81-44 \le k \le 81. The Intermediate Value Theorem alone would only have given “at least one”; the sign of p′p' is what turns it into “exactly one”.

Exercise 7: Monotonicity as a tool: one-to-one functions and inequalities

The corollary is a tool, not only a description. A function with f′>0f' > 0 on an interval is increasing there, hence one-to-one, hence invertible (chapter 2). And to prove F(x)>G(x)F(x) > G(x) for x>ax > a, study the difference d=F−Gd = F - G: if dd is continuous on [a,∞)[a, \infty), d′>0d' > 0 on (a,∞)(a, \infty) and d(a)=0d(a) = 0, then d(x)>d(a)=0d(x) > d(a) = 0 for x>ax > a.

The corollary needs continuity on the CLOSED interval, starting value included: the value d(a)d(a) is half of the proof.

  • a) Let f(x)=x3+3x−2f(x) = x^3 + 3x - 2. Show that ff is increasing on (−∞,∞)(-\infty, \infty), hence has an inverse, and compute (f−1)′(2)\left(f^{-1}\right)'(2).
  • b) Prove that ln⁡(1+x)>x1+x\ln(1 + x) > \frac{x}{1 + x} for every x>0x > 0.
  • c) Prove that 2x>3−1x2\sqrt x > 3 - \frac{1}{x} for every x>1x > 1.
  • d) A student proves b) in one line: “g′(x)>0g'(x) > 0 for x>0x > 0, so g(x)>0g(x) > 0 for x>0x > 0.” What is missing? Show with an example that the line proves nothing.

Type your answers, the page tells you right or wrong 0/7

a)
b)
c)
d)
Show the solution

Answers

  • a) f′(x)=3x2+3>0f'(x) = 3x^2 + 3 > 0: increasing, one-to-one; f(1)=2f(1) = 2, (f−1)′(2)=1f′(1)=16\left(f^{-1}\right)'(2) = \frac{1}{f'(1)} = \frac{1}{6}
  • b) g(x)=ln⁡(1+x)−x1+xg(x) = \ln(1 + x) - \frac{x}{1 + x}, g′(x)=x(1+x)2>0g'(x) = \frac{x}{(1 + x)^2} > 0 on (0,∞)(0, \infty), g(0)=0g(0) = 0
  • c) d(x)=2x+1x−3d(x) = 2\sqrt x + \frac{1}{x} - 3, d′(x)=x3/2−1x2>0d'(x) = \frac{x^{3/2} - 1}{x^2} > 0 for x>1x > 1, d(1)=0d(1) = 0
  • d) The starting value g(0)=0g(0) = 0 and the continuity of gg at 00; g−1g - 1 has the same derivative and g(x)−1<0g(x) - 1 < 0 for small x>0x > 0.

a) f′(x)=3x2+3=3(x2+1)f'(x) = 3x^2 + 3 = 3(x^2 + 1), and x2+1≥1x^2 + 1 \ge 1: f′(x)>0f'(x) > 0 for every xx. By the corollary, ff is increasing on every interval [a,b][a, b], hence on the whole line; so two different numbers have different images, ff is one-to-one and has an inverse f−1f^{-1}. To differentiate the inverse at 22 (chapter 12), first find which xx has f(x)=2f(x) = 2: x3+3x−2=2x^3 + 3x - 2 = 2 gives x3+3x−4=0x^3 + 3x - 4 = 0, with the obvious root x=1x = 1 (and no other, ff being one-to-one). So f−1(2)=1f^{-1}(2) = 1 and (f−1)′(2)=1f′(1)=16\left(f^{-1}\right)'(2) = \frac{1}{f'(1)} = \frac{1}{6}. Without the monotonicity, “the” xx with f(x)=2f(x) = 2 would not be known to be unique.

b) Let g(x)=ln⁡(1+x)−x1+xg(x) = \ln(1 + x) - \frac{x}{1 + x} on [0,∞)[0, \infty), continuous there. The derivative of x1+x\frac{x}{1 + x} is (1+x)−x(1+x)2=1(1+x)2\frac{(1 + x) - x}{(1 + x)^2} = \frac{1}{(1 + x)^2}. So g′(x)=11+x−1(1+x)2g'(x) = \frac{1}{1 + x} - \frac{1}{(1 + x)^2}, and the gesture is the common denominator (1+x)2(1 + x)^2: g′(x)=(1+x)−1(1+x)2=x(1+x)2g'(x) = \frac{(1 + x) - 1}{(1 + x)^2} = \frac{x}{(1 + x)^2}, which is >0> 0 for x>0x > 0. By the corollary, gg is increasing on [0,∞)[0, \infty), and for x>0x > 0, g(x)>g(0)=ln⁡1−0=0g(x) > g(0) = \ln 1 - 0 = 0. That is the inequality. Left as two separate fractions, the sign of g′g' is invisible.

c) Let d(x)=2x+1x−3d(x) = 2\sqrt x + \frac{1}{x} - 3 on [1,∞)[1, \infty), continuous there, with d(1)=2+1−3=0d(1) = 2 + 1 - 3 = 0. d′(x)=1x−1x2d'(x) = \frac{1}{\sqrt x} - \frac{1}{x^2}. Common denominator x2x^2, using 1x=x3/2x2\frac{1}{\sqrt x} = \frac{x^{3/2}}{x^2} (since x3/2⋅x−2=x−1/2x^{3/2} \cdot x^{-2} = x^{-1/2}): d′(x)=x3/2−1x2d'(x) = \frac{x^{3/2} - 1}{x^2}. For x>1x > 1, x3/2>1x^{3/2} > 1, so d′(x)>0d'(x) > 0. By the corollary, dd is increasing on [1,∞)[1, \infty), and d(x)>d(1)=0d(x) > d(1) = 0 for x>1x > 1: 2x+1x>32\sqrt x + \frac{1}{x} > 3, which is the inequality. The exponent rule x3/2⋅x−2=x−1/2x^{3/2} \cdot x^{-2} = x^{-1/2} is exactly the kind of step the department's algebra tutorials exist for: check it before relying on it.

d) A positive derivative says that gg goes UP, not that it is positive: an increasing function can be negative everywhere. Take k(x)=g(x)−1k(x) = g(x) - 1: k′(x)=g′(x)>0k'(x) = g'(x) > 0 on (0,∞)(0, \infty), yet k(x)k(x) is close to g(0)−1=−1g(0) - 1 = -1 for small x>0x > 0, so k(x)<0k(x) < 0 there. The one-line proof applies word for word to kk and “proves” a false statement. What is missing: gg is continuous on [0,∞)[0, \infty) (so the corollary covers the closed interval, point 00 included), and g(0)=0g(0) = 0. Then g(x)>g(0)=0g(x) > g(0) = 0. On a paper, the value at the starting point is written explicitly, or the proof earns part marks at best.

Exercise 8: Five statements to correct

Each statement below was written on a MATH 203 assignment, and each is false as written. Say what is wrong, give a counterexample, and write a correct statement.

  • a) If ff is differentiable and increasing on (a,b)(a, b), then f′(x)>0f'(x) > 0 for every xx in (a,b)(a, b).
  • b) If f′f' changes from positive to negative at cc, then ff has a local maximum at cc.
  • c) If ff is increasing on an interval II and f(x)≠0f(x) \ne 0 on II, then 1f\frac{1}{f} is increasing on II.
  • d) If f′(x)=x+1x−2f'(x) = \frac{x + 1}{x - 2}, then f′(x)>0f'(x) > 0 exactly when x+1>0x + 1 > 0, so ff is increasing on (−1,∞)(-1, \infty).
  • e) If f′(x)>0f'(x) > 0 for every xx in (0,1)(0, 1), then f(0)<f(1)f(0) < f(1).

Type your answers, the page tells you right or wrong 0/6

a)
b)
c)
d)
e)
Show the solution

Answers

  • a) False: (x−1)3(x - 1)^3 is increasing on (0,2)(0, 2) with f′(1)=0f'(1) = 0. True: f′(x)≥0f'(x) \ge 0 on (a,b)(a, b).
  • b) False: 1x2\frac{1}{x^2} at 00, not in the domain. True if cc is in the domain and ff is continuous at cc.
  • c) False: f(x)=xf(x) = x on (0,∞)(0, \infty), 1x\frac{1}{x} decreases. True: 1f\frac{1}{f} is DEcreasing, (1f)′=−f′f2\left(\frac{1}{f}\right)' = -\frac{f'}{f^2}.
  • d) False: f′(0)=−12f'(0) = -\frac{1}{2}. True: f′>0f' > 0 on (−∞,−1)(-\infty, -1) and on (2,∞)(2, \infty).
  • e) False without continuity at the ends: f(x)=xf(x) = x on [0,1)[0, 1), f(1)=−1f(1) = -1. True if ff is continuous on [0,1][0, 1].

a) FALSE. f(x)=(x−1)3f(x) = (x - 1)^3 is increasing on (0,2)(0, 2): f′(x)=3(x−1)2>0f'(x) = 3(x - 1)^2 > 0 except at 11, and a derivative zero at one isolated point does not stop the increase. Yet f′(1)=0f'(1) = 0, not >0> 0. The corollary goes in one direction only: f′>0f' > 0 implies increasing, not the converse. Correct statement: if ff is differentiable and increasing on (a,b)(a, b), then f′(x)≥0f'(x) \ge 0 for every xx in (a,b)(a, b), since every difference quotient f(x+h)−f(x)h\frac{f(x + h) - f(x)}{h} is positive and a limit of positive numbers is ≥0\ge 0.

b) FALSE. f(x)=1x2f(x) = \frac{1}{x^2} has f′(x)=−2x3f'(x) = -\frac{2}{x^3}, positive for x<0x < 0 and negative for x>0x > 0: the sign goes from ++ to −- at 00. But 00 is not in the domain, and near 00 the function grows without bound: no maximum. The First Derivative Test is stated at a critical point cc, a point OF THE DOMAIN where ff is continuous. Correct statement: if cc is in the domain, ff is continuous at cc and f′f' changes from positive to negative at cc, then ff has a local maximum at cc. On a sign chart, a number outside the domain gets a double bar and no verdict.

c) FALSE. f(x)=xf(x) = x is increasing and nonzero on (0,∞)(0, \infty), and 1f(x)=1x\frac{1}{f(x)} = \frac{1}{x} is decreasing there. If ff is differentiable, the derivative shows why: (1f)′=−f′f2\left(\frac{1}{f}\right)' = -\frac{f'}{f^2}, with f2>0f^2 > 0, so 1f\frac{1}{f} has the sign OPPOSITE to f′f'. Correct statement: if ff is increasing and never zero on an interval II, then 1f\frac{1}{f} is decreasing on II. The same holds without differentiability, since for u<vu < v with f(u)f(u) and f(v)f(v) of the same sign, 1f(u)>1f(v)\frac{1}{f(u)} > \frac{1}{f(v)}; on an interval, a continuous ff that is never zero keeps one sign.

d) FALSE. At x=0x = 0, x+1=1>0x + 1 = 1 > 0 but f′(0)=1−2=−12<0f'(0) = \frac{1}{-2} = -\frac{1}{2} < 0. The student multiplied the inequality x+1x−2>0\frac{x + 1}{x - 2} > 0 by x−2x - 2 without knowing its sign; for x<2x < 2 that multiplication reverses the inequality. The quotient is positive when numerator and denominator have the SAME sign: both positive for x>2x > 2, both negative for x<−1x < -1. Correct statement: f′>0f' > 0 on (−∞,−1)(-\infty, -1) and on (2,∞)(2, \infty), f′<0f' < 0 on (−1,2)(-1, 2), so ff is increasing on (−∞,−1](-\infty, -1] and on (2,∞)(2, \infty), separately. An inequality is multiplied only by a quantity of KNOWN sign; otherwise, a sign chart.

e) FALSE. Let f(x)=xf(x) = x for 0≤x<10 \le x < 1 and f(1)=−1f(1) = -1. Then f′(x)=1>0f'(x) = 1 > 0 on (0,1)(0, 1), yet f(1)=−1<0=f(0)f(1) = -1 < 0 = f(0). The corollary asks that ff be continuous on the CLOSED interval [0,1][0, 1], and this ff jumps at 11. Correct statement: if ff is continuous on [0,1][0, 1] and f′(x)>0f'(x) > 0 on (0,1)(0, 1), then ff is increasing on [0,1][0, 1], so f(0)<f(1)f(0) < f(1). Differentiability at the endpoints is NOT needed; continuity there is.

Exercise 9: Weekly sales after a launch: when do they rise, when do they fall?

A studio releases a video game. Its weekly sales, in thousands of copies per week, tt weeks after the launch, are modelled by S(t)=720tt2+36S(t) = \frac{720t}{t^2 + 36} for t≥0t \ge 0. The figure shows the graph of SS.

A sales report is written in words: “sales rise during the first ... weeks, peak at ... thousand copies per week, then fall.” The sign of S′S' writes that sentence, and the algebra of the quotient rule decides whether it is right.

51015202530354010203040506070y = S(t)t (weeks)S (thousand copies per week)
  • a) Show that S′(t)=720(6−t)(6+t)(t2+36)2S'(t) = \frac{720(6 - t)(6 + t)}{(t^2 + 36)^2}.
  • b) Write the sales report: when do weekly sales rise, when do they fall, which week is best, and what are the best weekly sales?
  • c) Compute S(3)S(3) and S(12)S(12). The marketing team says: sales were the same in week 33 and in week 1212, so nothing happened in between. Correct them.
  • d) During which period are weekly sales at least 3636 thousand copies? How many weeks is that?
  • e) Compute S′(3)S'(3) and S′(12)S'(12), with units, and interpret each in one sentence.

Type your answers, the page tells you right or wrong 0/10

a)
b)
Sales fall for tt in ,
c)
d)
Sales at least 3636 thousand for tt in ,
e)
Show the solution

Answers

  • a) Numerator 720(t2+36)−720t⋅2t=720(36−t2)720(t^2 + 36) - 720t \cdot 2t = 720(36 - t^2)
  • b) Rise on [0,6][0, 6], fall for t≥6t \ge 6; best week t=6t = 6, S(6)=60S(6) = 60 thousand copies per week
  • c) S(3)=S(12)=48S(3) = S(12) = 48; sales rose to the peak of 6060 at week 66, then fell back
  • d) 2≤t≤182 \le t \le 18, a period of 1616 weeks
  • e) S′(3)=9.6S'(3) = 9.6 and S′(12)=−2.4S'(12) = -2.4 thousand copies per week, per week

a) Quotient rule with u=720tu = 720t, v=t2+36v = t^2 + 36: S′(t)=720(t2+36)−720t⋅2t(t2+36)2S'(t) = \frac{720(t^2 + 36) - 720t \cdot 2t}{(t^2 + 36)^2}. Take out the common factor 720720 BEFORE expanding: the numerator is 720(t2+36−2t2)=720(36−t2)=720(6−t)(6+t)720\left(t^2 + 36 - 2t^2\right) = 720(36 - t^2) = 720(6 - t)(6 + t). Expanding first gives 720t2+25920−1440t2720t^2 + 25920 - 1440t^2, the same thing with three times more chances of an arithmetic slip.

b) For t≥0t \ge 0: 720>0720 > 0, 6+t>06 + t > 0 and (t2+36)2>0(t^2 + 36)^2 > 0, so the sign of S′S' is the sign of 6−t6 - t: positive for 0≤t<60 \le t < 6, negative for t>6t > 6. By the corollary, SS is increasing on [0,6][0, 6] and decreasing on [6,∞)[6, \infty), and by the First Derivative Test the local maximum is at t=6t = 6, with S(6)=432072=60S(6) = \frac{4320}{72} = 60. The report: weekly sales rise during the first six weeks, peak in week 66 at 6060 thousand copies per week, then decline week after week. Since SS increases up to 66 and decreases after, the peak is also the best week overall.

c) S(3)=216045=48S(3) = \frac{2160}{45} = 48 and S(12)=8640180=48S(12) = \frac{8640}{180} = 48: the same weekly sales. But SS is not constant between them: it increases on [3,6][3, 6] up to 6060, then decreases on [6,12][6, 12] back to 4848. Equal values at two times say nothing about what happens between; only the sign of S′S' does. There is even a pattern: S(36t)=720⋅36/t362/t2+36=720t36+t2=S(t)S\left(\frac{36}{t}\right) = \frac{720 \cdot 36/t}{36^2/t^2 + 36} = \frac{720t}{36 + t^2} = S(t), so every weekly level below the peak is reached twice, once on the way up (before week 66) and once on the way down, at times whose product is 3636.

d) Solve 720tt2+36≥36\frac{720t}{t^2 + 36} \ge 36. Multiplying by t2+36t^2 + 36 is allowed here because it is POSITIVE, so the inequality keeps its direction: 720t≥36t2+1296720t \ge 36t^2 + 1296, divide by 3636: 20t≥t2+3620t \ge t^2 + 36, that is t2−20t+36≤0t^2 - 20t + 36 \le 0, (t−2)(t−18)≤0(t - 2)(t - 18) \le 0, so 2≤t≤182 \le t \le 18. Weekly sales are at least 3636 thousand from week 22 to week 1818, a period of 1616 weeks. Consistent with c): 2×18=362 \times 18 = 36, and the peak week 66 lies inside.

e) S′(3)=720(3)(9)452=194402025=9.6S'(3) = \frac{720(3)(9)}{45^2} = \frac{19440}{2025} = 9.6 and S′(12)=720(−6)(18)1802=−7776032400=−2.4S'(12) = \frac{720(-6)(18)}{180^2} = \frac{-77760}{32400} = -2.4, in thousand copies per week, per week. In week 33, weekly sales are growing at about 96009600 copies per week each week; in week 1212 they are shrinking by about 24002400 copies per week each week. The same level of sales, 4848 thousand, hides two opposite movements: the sign of S′S' is what the report must say.

Exercise 10: A final exam question: a logarithmic family and the parameter that makes the minimum touch zero

For a real constant aa, let fa(x)=x2−aln⁡xf_a(x) = x^2 - a\ln x on (0,∞)(0, \infty). The figure shows faf_a for a=2a = 2, a=2ea = 2e and a=8a = 8, with the lowest point of each curve.

The algebra of the chapter in its logarithmic version: one fraction over xx, the domain x>0x > 0 used to drop factors and reject roots, and the laws of logarithms to simplify the values.

0.511.522.533.5-2-11234567green: a = 2orange: a = 2eblue: a = 8x
  • a) For a=8a = 8: write f8′(x)f_8'(x) as one factored fraction, find the intervals of increase and decrease and the local extreme value of f8f_8, exactly.
  • b) Show that if a≤0a \le 0, faf_a is increasing on (0,∞)(0, \infty) and has no local extremum.
  • c) Let a>0a > 0. Show that faf_a has exactly one local extremum, a minimum at x=a2x = \sqrt{\frac{a}{2}}, with value m(a)=a2(1−ln⁡a2)m(a) = \frac{a}{2}\left(1 - \ln\frac{a}{2}\right).
  • d) Find the value of a>0a > 0 for which the minimum value is 00, and deduce an inequality valid for every x>0x > 0.
  • e) Study the function mm on (0,∞)(0, \infty) with the First Derivative Test. For which aa is the minimum value m(a)m(a) the largest, and what is it?

Type your answers, the page tells you right or wrong 0/9

a)
b)
c)
d)
e)
Show the solution

Answers

  • a) f8′(x)=2(x−2)(x+2)xf_8'(x) = \frac{2(x - 2)(x + 2)}{x}: decreasing on (0,2](0, 2], increasing on [2,∞)[2, \infty); local min f8(2)=4−8ln⁡2f_8(2) = 4 - 8\ln 2
  • b) fa′(x)=2x2−ax>0f_a'(x) = \frac{2x^2 - a}{x} > 0 on (0,∞)(0, \infty) when a≤0a \le 0
  • c) fa′(x)=2(x−a/2)(x+a/2)xf_a'(x) = \frac{2\left(x - \sqrt{a/2}\right)\left(x + \sqrt{a/2}\right)}{x}, −- then ++; m(a)=a2−a2ln⁡a2m(a) = \frac{a}{2} - \frac{a}{2}\ln\frac{a}{2}
  • d) a=2ea = 2e; x2≥2eln⁡xx^2 \ge 2e\ln x for every x>0x > 0, with equality only at x=ex = \sqrt e
  • e) m′(a)=−12ln⁡a2m'(a) = -\frac{1}{2}\ln\frac{a}{2}: mm increases on (0,2](0, 2], decreases on [2,∞)[2, \infty); largest value m(2)=1m(2) = 1

a) f8′(x)=2x−8xf_8'(x) = 2x - \frac{8}{x}. One fraction over xx: f8′(x)=2x2−8x=2(x2−4)x=2(x−2)(x+2)xf_8'(x) = \frac{2x^2 - 8}{x} = \frac{2(x^2 - 4)}{x} = \frac{2(x - 2)(x + 2)}{x}. On the domain x>0x > 0, both xx and x+2x + 2 are positive: the sign of f8′f_8' is the sign of x−2x - 2. The root −2-2 is not in the domain and is rejected, not tested. So f8f_8 is decreasing on (0,2](0, 2] and increasing on [2,∞)[2, \infty), with a local minimum f8(2)=4−8ln⁡2f_8(2) = 4 - 8\ln 2, about −1.545-1.545, the lowest point of the blue curve.

b) fa′(x)=2x−ax=2x2−axf_a'(x) = 2x - \frac{a}{x} = \frac{2x^2 - a}{x}. If a≤0a \le 0, then −a≥0-a \ge 0 and 2x2−a≥2x2>02x^2 - a \ge 2x^2 > 0 for x>0x > 0, while the denominator xx is positive: fa′(x)>0f_a'(x) > 0 on (0,∞)(0, \infty). By the corollary faf_a is increasing on every closed interval inside (0,∞)(0, \infty), hence on (0,∞)(0, \infty), and without a sign change there is no local extremum. For a=0a = 0, f0(x)=x2f_0(x) = x^2 on (0,∞)(0, \infty): its famous minimum at 00 is outside the domain.

c) For a>0a > 0, 2x2−a=2(x2−a2)=2(x−a2)(x+a2)2x^2 - a = 2\left(x^2 - \frac{a}{2}\right) = 2\left(x - \sqrt{\frac{a}{2}}\right)\left(x + \sqrt{\frac{a}{2}}\right). On (0,∞)(0, \infty) the factors x+a/2x + \sqrt{a/2} and xx are positive, so fa′f_a' has the sign of x−a/2x - \sqrt{a/2}: negative, then positive. The only critical point in the domain is a/2\sqrt{a/2}, and the First Derivative Test gives a local minimum there, the only local extremum. Its value: fa(a/2)=a2−aln⁡a2f_a\left(\sqrt{a/2}\right) = \frac{a}{2} - a\ln\sqrt{\frac{a}{2}}, and the law of logarithms ln⁡u=12ln⁡u\ln\sqrt u = \frac{1}{2}\ln u turns it into a2−a2ln⁡a2=a2(1−ln⁡a2)\frac{a}{2} - \frac{a}{2}\ln\frac{a}{2} = \frac{a}{2}\left(1 - \ln\frac{a}{2}\right). Check with a=8a = 8: 4(1−ln⁡4)=4−8ln⁡24(1 - \ln 4) = 4 - 8\ln 2, as in a).

d) Since a>0a > 0, m(a)=0m(a) = 0 exactly when 1−ln⁡a2=01 - \ln\frac{a}{2} = 0, that is a2=e\frac{a}{2} = e, a=2ea = 2e. For this value, f2ef_{2e} is decreasing on (0,e]\left(0, \sqrt e\right] and increasing on [e,∞)\left[\sqrt e, \infty\right) (since a/2=e\sqrt{a/2} = \sqrt e), so f2e(x)≥f2e(e)=0f_{2e}(x) \ge f_{2e}(\sqrt e) = 0 for every x>0x > 0, with equality only at e\sqrt e: that is, x2≥2eln⁡xx^2 \ge 2e\ln x for every x>0x > 0. On the figure, the orange curve touches the axis at (e,0)\left(\sqrt e, 0\right).

e) m(a)=a2−a2ln⁡a2m(a) = \frac{a}{2} - \frac{a}{2}\ln\frac{a}{2}. Product rule on the second term, with (ln⁡a2)′=1a\left(\ln\frac{a}{2}\right)' = \frac{1}{a}: m′(a)=12−(12ln⁡a2+a2⋅1a)=−12ln⁡a2m'(a) = \frac{1}{2} - \left(\frac{1}{2}\ln\frac{a}{2} + \frac{a}{2} \cdot \frac{1}{a}\right) = -\frac{1}{2}\ln\frac{a}{2}. It is positive when a2<1\frac{a}{2} < 1 and negative when a2>1\frac{a}{2} > 1: mm is increasing on (0,2](0, 2] and decreasing on [2,∞)[2, \infty), so its largest value is m(2)=1⋅(1−ln⁡1)=1m(2) = 1 \cdot (1 - \ln 1) = 1. The green curve of the figure, a=2a = 2, has the highest lowest point of the whole family, (1,1)(1, 1). The First Derivative Test has just been applied to a function of the PARAMETER, which is the same test on a different variable.

© Ahmed Squalli Houssaini. Exercise set published at www.letuteurscientifique.ca/en/exercices/math203-monotonic-functions. Free for personal and classroom use; republishing it elsewhere requires written permission (legal notice).

See also

Struggling with MATH 203?

I tutor first-year calculus at Concordia and McGill, in English or in French, in Montreal or online, algebra gaps included. Get in touch for a first session.

Site by Studio Squalli