Just as you can add and subtract functions, you can multiply and divide them. Products and quotients build some of the most useful models there are: revenue is price times quantity, and a bouncing spring that slowly settles down is an exponential function times a sinusoidal one. Trig functions here use radians.
Put the same x into both functions, then multiply or divide the outputs. Careful with notation: (fg)(x) means the productf(x)⋅g(x). It is not the same as f(g(x)), which is a composition.
fg: the overlap of the domains of f and g (just like f+g).
gf: the overlap of the domains, and also remove every x where g(x)=0.
Find the domain before you simplify. If a factor cancels, its zero is still excluded, and the graph has a hole there instead of an asymptote. This is the same idea as the restrictions on rational expressions.
Think about the y-values of the two pieces at each x:
Zeros:(fg)(x)=0 exactly when f(x)=0 or g(x)=0 (inside the common domain). So the product’s zeros are the zeros of both factors put together.
Sign: the product is positive where the factors have the same sign, and negative where they have opposite signs.
Where a factor equals 1: if g(x)=1, then (fg)(x)=f(x), so the graphs meet there. If g(x)=−1, the product is −f(x).
Envelopes: since −1≤sinx≤1, the product f(x)sinx always stays between −f(x) and f(x) (when f(x)≥0). The curves y=±f(x) form an envelope that the product bounces between.
A spring, a pendulum, or a guitar string vibrates back and forth, but friction makes each swing a little smaller. That is a sinusoidal function multiplied by a decreasing exponential:
d(t)=Abtcos(kt)(0<b<1)
The cosine supplies the back-and-forth motion, and Abt is the envelope that shrinks the amplitude toward 0.
The rules work like the signs in multiplication (think of even as + and odd as −):
f
g
fg and gf
even
even
even
odd
odd
even
even
odd
odd
For example, if f and g are both odd, then (fg)(−x)=f(−x)g(−x)=(−f(x))(−g(x))=f(x)g(x), so fg is even. And tanx=cosxsinx is odd divided by even, which is odd.
Let f(x)=x2−9 and g(x)=x−3. Find (gf)(x), state its domain, and describe its graph.
Solution. First the domain: g(x)=0 at x=3, so x=3 is excluded. Now simplify:
(gf)(x)=x−3x2−9=x−3(x−3)(x+3)=x+3,x=3
The domain is {x∈R∣x=3}. The graph is the line y=x+3 with a hole at x=3. The hole’s y-coordinate comes from the simplified form: 3+3=6, so the hole is at (3,6).
y=x−3x2−9 is the line y=x+3 with the point (3,6) missing.
Compare (fg)(x)=(x−3)(x+3)x−3=x+31 for x=±3. Here the factor x+3 does not cancel, so x=−3 gives a vertical asymptote, while x=3 gives a hole at (3,61).
A weight on a spring is pulled down and released. Its displacement below its resting position (positive means below), in centimetres, after t seconds is modelled by
d(t)=12(0.8)tcos(πt)
(a) Find d(0), d(1), d(2) and d(3), and describe the motion.
(b) Find the times in the first 4 s when the weight passes through its resting position.
(c) After how many seconds is the envelope less than 1 cm?
Solution.
(a) cos(πt) is 1 at even whole numbers and −1 at odd whole numbers:
The weight swings from one side to the other every second, and each swing is 80% as big as the one before. The cosine makes it oscillate with period π2π=2 s, and the factor 12(0.8)t shrinks the amplitude.
The product bounces between the envelope curves d=±12(0.8)t.
(b) d(t)=0 when either factor is 0. The factor 12(0.8)t is never 0, so we need cos(πt)=0:
πt=2π,23π,25π,27π⇒t=0.5,1.5,2.5,3.5
The weight passes through its resting position at 0.5, 1.5, 2.5 and 3.5 s.
Finding the domain after simplifying.x−3x2−9 simplifies to x+3, but x=3 is still excluded. Always find the restrictions from the original quotient.
Confusing fg with f(g(x)).(fg)(x) multiplies the outputs; f(g(x)) puts one function inside the other. For f(x)=x+1 and g(x)=x2, (fg)(x)=x3+x2 but f(g(x))=x2+1.
Expecting zeros from an exponential factor. In 12(0.8)tcos(πt), only the cosine can be 0: an exponential like bt is always positive.
Saying “odd × odd = odd”. The rule follows signs: two negatives make a positive, so odd × odd is even. Check with x⋅x=x2.
Forgetting the zeros of g are excluded even when f is also zero there. At x=3 in Example 2, both f and g are 0, and 00 is undefined. That’s a hole, not a point on the graph.
Flipping the wrong way when dividing by a negative log.log0.8<0, so dividing by it reverses the inequality in Example 3(c).
2. (Warm-up) If f(3)=−2 and g(3)=5, find (fg)(3), (gf)(3) and (fg)(3).
Solution
(fg)(3)=(−2)(5)=−10
(gf)(3)=5−2=−0.4
(fg)(3)=−25=−2.5
3. (Warm-up) Use the table to find (fg)(x) and (gf)(x) at each x. Where is gf undefined?
x
0
1
2
3
f(x)
4
−3
6
0
g(x)
2
3
0
−5
Solution
x
0
1
2
3
(fg)(x)
8
−9
0
0
(gf)(x)
2
−1
undefined
0
gf is undefined at x=2, because g(2)=0. The product is 0 at x=2 and x=3, where one of the factors is 0.
4. (Core) Let f(x)=x2−x−6 and g(x)=x+2. Simplify (gf)(x), state its domain, and give the coordinates of any hole.
Solution
g(x)=0 at x=−2, so that’s excluded. Factor and simplify:
(gf)(x)=x+2(x−3)(x+2)=x−3,x=−2
Domain: {x∈R∣x=−2}. The graph is the line y=x−3 with a hole at x=−2, where y=−2−3=−5: the hole is at (−2,−5).
5. (Core) Let f(x)=x+1 and g(x)=x2−4. State the domain of gf.
Solution
f needs x≥−1. g is defined everywhere, but g(x)=0 at x=±2. The value x=−2 is already outside x≥−1, so only x=2 needs to be removed:
{x∈R∣x≥−1,x=2}
6. (Core) Let h(x)=(x−1)⋅2x. Without graphing technology, find the zeros and the y-intercept of h, and state where h(x) is positive and where it is negative.
Solution
2x is always positive, so it has no zeros. The only zero comes from x−1=0: x=1.
y-intercept: h(0)=(0−1)(1)=−1.
Sign: since 2x>0, h(x) has the same sign as x−1. So h(x)<0 for x<1 and h(x)>0 for x>1.
7. (Core) A café sells 200 muffins a day at $3.00 each. For every $0.10 decrease in price, it sells 20 more muffins. After x price decreases, the price is p(x)=3−0.1x dollars and the number sold is q(x)=200+20x.
(a) Find the revenue function R(x)=(pq)(x).
(b) What price gives the maximum revenue, and what is that revenue?
(b) sin2x=(sinx)(sinx) is odd × odd: g(−x)=(sin(−x))2=(−sinx)2=sin2x=g(x). Even.
(c) tanx=cosxsinx is odd ÷ even. Odd.
General rule: suppose f is odd and g is even. Then, wherever g(x)=0,
(gf)(−x)=g(−x)f(−x)=g(x)−f(x)=−(gf)(x)
so gf is odd.
9. (Challenge) Water drains from a storm sewer into a lake. After t seconds, the concentration of a pollutant in the water is c(t)=2t2 kilograms per cubic metre, and the water flows out at w(t)=t4+163 cubic metres per second.
(a) Show that the rate at which the pollutant enters the lake, in kilograms per second, is r(t)=t4+166t2.
(b) Make a table of r(t) for t=1,1.5,2,2.5,3 (to 2 decimal places).
(c) Show that r(t)≤0.75 for every t>0, so the maximum rate is 0.75 kg/s at t=2 s. (Hint: (t2−4)2≥0.)
Solution
(a) (kg per m³) × (m³ per s) = kg per s, so the rate is the product:
r(t)=c(t)w(t)=2t2⋅t4+163=t4+166t2
(b)
t (s)
1
1.5
2
2.5
3
r(t) (kg/s)
0.35
0.64
0.75
0.68
0.56
The table suggests a maximum near t=2.
(c) Expand the square:
(t2−4)2t4−8t2+16t4+16≥0≥0≥8t2
Since t4+16≥8t2>0, a bigger denominator gives a smaller fraction:
r(t)=t4+166t2≤8t26t2=0.75
Equality happens only when t2−4=0, so at t=2. The maximum rate is 0.75 kg/s, at t=2 s. (Graphing technology shows the same peak.)