MATH-341: FALL SEMESTER 2017 HOMEWORK 12 CORRESPONDING QUIZ ON 12/6 ANSWER KEY The following questions cover R Section 6.4: 1. Suppose that the pdf of a random variable X is: f (x) = 3x2 1{0<x<1} . Let Y = 1 − X 2 . Find the pdf of Y . First, we have: FY (y) = Pr(Y ≤ y) = Pr(1 − X 2 ≤ y) = Pr(1 − y ≤ X 2 ) 1 − Pr(1 − y > X 2 ) p p = 1 − Pr( 1 − y > X > − 1 − y) p = 1 − Pr(X < 1 − y) = where the last line follows from the support of X: 0 < X < 1. From this, we have: fY (y) d FY (y) dy p d = [1 − Pr(X < 1 − y)] dy "Z √ # 1−y d 2 3x dx = − dy 0 # " √ 1−y d = − x3 dy 0 = d [(1 − y)3/2 ] dy 3p 1−y 2 = − = Also, since Y = 1 − X 2 , this tells us that the support of Y must be 0 < Y < 1. Therefore, 3p 1 − y · 1{0<y<1} fY (y) = 2 2. Suppose that a random variable X can have each of the seven values -3,-2,-1,0,1,2,3 with equal probability. Determine the pmf of Y = X 2 − X. First we note that: 0 for x = 0 or x = 1 2 for x = −1 or x = 2 y= 6 for x = −2 or x = 3 12 for x = −3 1 Thus, Pr(Y = 0) = Pr(Y = 2) = Pr(Y = 6) = Pr(Y = 12) = 2 7 2 7 2 7 1 7 3. Suppose that the pdf of X is: f (x) = 1 x · 1{0<x<2} 2 Determine the pdf of Y = 4 − X 3 . FY (y) = Pr(Y ≤ y) = Pr(4 − X 3 ≤ y) Pr(4 − y ≤ X 3 ) 1 = Pr (4 − y) 3 ≤ X Z 2 1 = x dx 1 (4−y) 3 2 = = x2 4 2 1 (4−y) 3 2 = ⇒ fY (y) = = = 1− (4 − y) 3 4 d FY (y) dy " # 2 d (4 − y) 3 1− dy 4 1 1 (4 − y)− 3 6 Also, the support of Y is −4 < Y < 4, so we have: fY (y) = 1 1 (4 − y)− 3 1{−4<y<4} 6 4. Suppose that the pdf of X is: f (x) = e−x 1{x>0} 2 Determine the pdf of Y = X 1/2 . FY (y) = = Pr(Y ≤ y) 1 Pr X 2 ≤ y Pr(X ≤ y 2 ) Z y2 = e−x dx = 0 y2 = −e −x 0 2 1 − e−y i 2 d h 1 − e−y = dy = ⇒ fY (y) = 2ye−y 2 Coming from the support of X: X > 0, we thus know that the support of Y is Y > 0, so: fY (y) = 2 2ye−y 1{y>0} 5. Let X have the uniform distribution on the interval [a, b], and let c be some constant such that c > 0. Let Y = cX + d where d is also a constant. Show that Y also has a uniform distribution, specifically on the interval [ca + d, cb + d]. If X is uniform over the interval [a, b], then: fX (x) = 1 1{a≤x≤b} b−a Now, let Y = cX + d. This function of X is monotonic, and note that: Y = cX + d Y − d = cX Y −d = X c y−d ⇒ h−1 (y) = c 3 so we can use: fY (y) = = = = = = fX [h−1 (x)] dh−1 dy 1 dh−1 1{a≤h−1 (x)≤b} b−a dy d y−d 1 × 1 y−d b − a {a≤ c ≤b} dy c 1 1 1{ca+d≤y≤cb+d} × b−a c 1 1{ca+d≤y≤cb+d} cb − ca 1 1{ca+d≤y≤cb+d} (cb + d) − (ca + d) where this is the pdf of a uniform random variable between ca + d and cb + d as required. 6. Suppose that you would like to simulate a variable X with the following pdf: f (x) = 1 (2 − x)1{0<x<2} 2 Show how you can use the Probability Integral Transformation (as shown in class, and a specific one shown in Example 6.5 on page 306), to transform a uniform random variable into one with this pdf. First, we find the cdf of X: Z FX (t) t = −∞ t Z 1 (2 − x)1{0<x<2} dx 2 1 (2 − x) dx 2 = 0 = x− x2 4 t 0 t2 = t− 4 Now, using the Probability Integral Transform, we have Y = FX (x), where we showed that this Y has a Uniform[0,1] distribution. −1 Therefore, if Y = FX (x), then we note that X = FX (y). So, we take the cdf of X and solve 4 for x: y −4y 4 − 4y 4 − 4y p ± 4 − 4y p − 4 − 4y p 2−2 1−y x2 4 = x2 − 4x = x− = x2 − 4x + 4 = (x − 2)2 = (x − 2) = x − 2 (since 0 < x < 2, so x − 2 must be negative) = x So, if we can simulate values of the variable Y which come from a Uniform distribution over the interval [0, 1], then we would just take these values, and plug them into the expression √ 2 − 2 1 − y, which will then give us values of x that will have the distribution of the variable X given by the pdf fX above. 7. Read and understand each question on the In-class activity from 11/27/17 See class notes. 8. Read and understand each question on the In-class activity from 12/1/17 See class notes. 5