We will continue our discussion of curve sketching. First, we will complete the example from class. I’d like to remind everyone that we showed that f ′ (x) is negative everywhere it is defined, and that limx→0 f ′ (x) = −∞. Furthermore, we computed: f ′′ (x) = (2/3x−5/3 )(20x4 + 17x2 − 1) 9(x2 − 1)3 So we are interested in determining when 20x4 + 17x2 − 1 = 0. To do this, make the substitution u = x2 to get: 20u2 + 17u − 1 = 0. Then the solution is √ −17 ± 369 u= 40 √ by the quadratic formula. Notice that only −17+40 369 is positive, which is important because u = x2 must be nonnegative. Therefore, we get that f ′′ (x) = 0 when x is equal to s √ −17 + 369 ± . 40 As usual, the next stepr is to construct a sign chart for thersecond derivative: r r region f ′′ (x) f (x) x < −1 √ −17 369 40 positive CCU −1 < x < − negative CCD − √ −17 369 < x < 0 40 negative CCD √ −17 369 40 positive CCU 0 < x < √ −17 369 < x < 1 40 negative CCD This tells us everything we need to know to complete the sketch. Notice that the concavity changes at x = 0 so we have an inflection point. However, it’s not a typical inflection point- there is a vertical tangent line at x = 0. Let’s move on to the next example I had planned for Friday. Problem 0.1. Sketch the graph of g(x) = x ln x. First, we note the domain: the expression is defined for x > 0. Furthermore, it is clear that limx→∞ g(x) = ∞. Let’s consider the first derivative: g ′ (x) = ln x + x · 1 x which is equal to ln x + 1. This is equal to zero when ln x = −1, which happens when x = 1e . Note that g ′ (x) < 0 for 0 < x < 1e and g ′ (x) > 0 for x > 1e . Now, we take a second derivative: 1 g ′′ (x) = x which is positive on (0, ∞). Therefore, there are no inflection points. The only question that remains is the behavior as x → 0. We need to evaluate: lim x ln x x→0 1 x > 1 positive CCU This appears to come out to 0 × ∞ which doesn’t mean anything. It would be nice if we could find a way to use L’Hôpital’s rule to evaluate this limit. 1 The trick is to rewrite x as 1/x . So we get ln x lim x→0 1/x Which is, by L’Hôpital’s rule: 1/x x→0 −1/x2 lim which is −x2 x→0 x lim which comes out to lim −x = 0 x→0 so the limit is zero. Problem 0.2. Sketch the graph of f= p x2 − x4 . The first thing to notice about this function is the restriction on the domain. We need x2 − x4 ≥ 0, which implies that −1 ≤ x ≤ 1. There is no need to talk about horizontal asymptotes for f (x) because the domain is bounded. Furthermore, f (x) is continuous on [−1, 1] so there are no vertical asymptotes either. Let’s take a derivative: 1 (2x − 4x3 )(x2 − x4 )−1/2 . 2 This is undefined at x = 0 as well as at x = ±1 because of the zero in the denominator. Let’s compute the left and right limits of f ′ (x) as x → 0. We could use L’Hôpital’s rule, but it’s easier just to factor: f ′ (x) = x − 2x3 lim √ x→0+ x2 − x4 x(1 − 2x2 ) = lim √ x→0+ x 1 − x2 1 − 2x2 = lim √ x→0+ 1 − x2 =1 2 We can also compute the left limit: x − 2x3 lim √ x→0− x2 − x4 x(1 − 2x2 ) √ = lim− x→0 −x 1 − x2 1 − 2x2 = lim− √ x→0 − 1 − x2 = −1 Note that these limits are different. If we take the limit as x → 1, we see that the numerator goes to −1 and the denominator goes to 0 from the right, implying that the limit is −∞. Similarly, looking in the region where x → −1, we see that the limit is ∞. So we expect the graph to be near vertical close to x = 1 and x = −1. Finally, we notice that the derivative is zero whenever 1 − 2x2 = 0, or √ 1 2 2 whenever x = 2 . This happens at ± 2 . The sign chart for f ′ (x) looks like this: √ √ √ √ region −1 < x < − 2 − 2 < x < 0 0 < x < 2 2<x<1 f ′ (x) positive negative positive negative f (x) increasing decreasing increasing decreasing The sign chart can be completed using the limit calculations from before (or √ 2 by plugging in test points). So f (x) has local maxima when x = ± 2 and a local minimum at x = 0. Now, let’s look at the second derivative: 1 f ′′ (x) = (1 − 6x2 )(x2 − x4 )−1/2 − (x − 2x3 )(x2 − x4 )−3/2 (2x − 4x3 ) 2 which simplifies to f ′′ (x) = (1 − 6x2 )(x2 − x4 )−1/2 − (x − 2x3 )2 (x2 − x4 )−3/2 This is again undefined at −1, 0, and 1. Let’s factor out a copy of (x2 − x4 )−3/2 : f ′′ (x) = (x2 − x4 )−3/2 ((x2 − x4 )(1 − 6x2 ) − (x − 2x3 )2 ) Notice that we can also factor an x2 out: f ′′ (x) = x2 (x2 − x4 )−3/2 ((1 − x2 )(1 − 6x2 ) − (1 − 2x2 )2 ) Now we need to bite the bullet and multiply things out: f ′′ (x) = x2 (x2 − x4 )−3/2 (1 − 7x2 + 6x4 − 1 + 4x2 − 4x4 ) this is f ′′ (x) = x2 (x2 − x4 )−3/2 (−3x2 + 2x4 ) 3 Notice that this is always negative which means that f (x) is concave down on (−1, 0) and (0, 1). Furthermore, notice the behavior near x = 0: The numerator has essentially 4 powers of x near zero and the denominator looks like (x2 )3/2 , which is x3 . So we expect the second derivative to approach 0 as x → 0. This is indeed what happens. Therefore, the function f ′′ (x) should be almost straight near x = 0. Putting this information together gives us the graph. 4