MATH 10B DISCUSSION SECTION PROBLEMS 3/21 – SOLUTIONS

JAMES ROWAN

1. Continuous random variables: pdfs and cdfs

(1) True/false practice:
(a) The function

0
 x≤0
π
F (x) = sin x 0<x< 2
x ≥ π2

1


is a CDF for some contiuous random variable X.

True . This function is right continuous (in fact, it is continuous), nondecreasing, and satisfies
limx→−∞ F (x) = 0, limx→∞ F (x) = 1, so it is a CDF.

(b) We require CDFs to be continuous, but we don’t require PDFs to be continuous.

False . We do not require CDFs to be continuous, we only require them to be right continuous. As
an example, the CDF of a Bernoulli random variable is only right-continuous, since for any negative
x, P (X ≤ x) = 0, but P (X ≤ 0) = (1 − p) and P (X ≤ x) = (1 − p) for all 0 ≤ x < 1.

(c) If f (x) is a PDF for a continuous random variable, then any antiderivative of f (x) is a CDF for that
random variable.

False . Only one antiderivative of f (x) will be a CDF for that random variable. For example, for the
random variable with PDF f (x) = 0 for x < 0, f (x) = e−x for x ≥ 0, the antiderivative is F (x) = C1
for x < 0, F (x) = C2 − e−x for x ≥ 0 (the different constants of integration come from the fact that
we can pick a different constant of integration when we integrate the two pieces of the PDF). In order
for limx→−∞ F (x) = 0 and limx→∞ F (x) = 1, we need to pick the constants of integration C1 and C2
so that C1 = 0 and C2 = 1; no other choice of constants of integration will work.
(2) (original) If the answer to T/F (a) is true, find the PDF for this random variable. If the answer to T/F
(a) is false, change the function in some way to make a CDF.

The answer T/F (a) is true, so we find the PDF for this random variable. We find the PDF from the
CDF by differentiating the CDF. Differentiating each of the pieces of F (x), we find that

0
 x≤0
f (x) = cos x 0 < x < π2
x ≥ π2

0


is the PDF for this random variable.

Note that we could change the x ≤ 0, 0 < x < π2 in the above to x < 0, 0 ≤ x < π2 without causing
any problems; since the PDF is related to probabilities only via integration and redefining our function at
a single point won’t change its integral, both PDFs will correspond to the same random variable.

(3) (original) Let


0
 x<0
π
f (x) = c cos x 0≤x≤ 3
x > π3 .

0

1
2 JAMES ROWAN

Find the value of the constant c such that f (x) is a pdf, then find the corresponding CDF and sketch it.

In order for f (x) to be a PDF, the total area under the graph of f (x) needs to be 1. Since f (x) is equal
R π/3
to 0 outside [0, π3 ], it is enough to find the value of c such that 0 c cos xdx = 1. We have

Z π/3
π/3
c cos xdx = c sin x 0
0
π
= c sin − c sin(0)
√ 3
3
=c· ,
2
√
so we need to choose c = 2 3 3 .
A sketch of the graph of the CDF is below:

(4) (Stewart/Day 12.5.9) Let f (x) = 1 for 0 ≤ x ≤
1, f (x) = 0 otherwise. Verify that f (x) is a PDF, find
the CDF, and use the CDF to find P 31 ≤ X ≤ 23 .

We see that f (x) is piecewise continuous, nonnegative everywhere, and that the area under the graph
R1
of f (x) is 0 1dx = 1 (outside the interval [0, 1], f (x) is identically zero so no area gets contributed except
between 0 and 1). This means f (x) is a PDF.
We find the CDF by integrating f (x). Since the integral of 1 is x + C and picking C = 0 gives us the
right limiting behavior (and makes F (x) the area-so-far function for f (x)), our CDF is


0
 x<0
F (x) = x 0≤x≤1

1 1 < x.


To find P 13 ≤ X ≤ 23 , we use that P 13 ≤ X ≤ 32 = F 23 − F 13 = 23 − 31 = 13 .

(5) (Stewart/Day 12.5.15) Let f (x) = 21 e−|x| for x ∈ R. Verify that f (x) is a PDF, find the CDF, and use
the CDF to find P (−5 ≤ X ≤ 10).
 MATH 10B DISCUSSION SECTION PROBLEMS 3/21 – SOLUTIONS 3

We see that f (x) is piecewise continuous and nonnegative everywhere. The area under the graph of f (x)
is
Z ∞ Z t
1 −|x| 1 −|x|
e dx = lim e dx
−∞ 2 t→∞ −t 2
Zt Z 0
1 −x 1 x
= lim e dx + e dx
t→∞ 0 2 −t 2
1 1 1 1
= lim − e−t + e−0 + e0 − e−t
t→∞ 2 2 2 2
= lim 1 − e−t = 1.
t→∞

Since we have checked all the requirements for having a PDF, this f (x) is a PDF.
To find the CDF, we use the area-so-far function for this PDF. On (−∞, 0), the antiderivative of
f (x) = 12 ex is F (x) = 12 ex +C; picking C = 0 ensures that limx→−∞ F (x) = 0. On [0, ∞), the antiderivative
of f (x) = 21 e−x is F (x) = C − 12 e−x ; picking C = 1 ensures that limx→∞ F (x) = 1. The CDF is thus the
piecewise-defined function (
1 x
e x<0
F (x) = 2 1 −x
1 − 2e x ≥ 0.
We have
   
1 −10 1 −5 1 −10
+ e−5 .


P (−5 ≤ X ≤ 10) = F (10) − F (−5) = 1 − e − e =1− e
2 2 2
2
(6) (adapted from Stewart/Day 12.5.17) Let f (x) = for x ≥ 0 and f (x) = 0 otherwise. Verify
π(1+x2 )
 √ 
that f (x) is a PDF and find the CDF. Find the probabilities P 0 ≤ X ≤ 33 and P (X ≥ 1).

We see that f (x) is piecewise continuous and nonnegative everywhere. Since f (x) is 0 for x < 0, the
area under the graph of f (x) is
Z ∞ Z t
2 2
2
dx = lim dx
0 π(1 + x ) t→∞ 0 π(1 + x2 )
2 t
= lim tan−1 (x) 0
t→∞ π
= 1 − 0 = 1.
Since we have checked all the requirements for having a PDF, this f (x) is a PDF.
To find the CDF, we use the area-so-far function for this PDF. For x < 0, the CDF F (x) will be 0, as
f (x) is zero for x less than zero. For x ≥ 0, we have
Z x Z x
2 2
F (x) = f (t)dt = 2
dt = tan−1 (x).
0 0 π(1 + t ) π
The CDF is thus the piecewise-defined function
(
0 x<0
F (x) = 2 −1
π tan (x) x ≥ 0.
 √  √ √ 
The probability P 0 ≤ X ≤ 33 is given by F ( 33 ) − F (0), which is 2
π · tan−1 3
3
= 2
π · π
6 = 1
3.
The probability P (X ≥ 1) is equal to 1 − P (X < 1), which is 1 − F (1) (since the CDF is the CDF for a
continuous random variable, P (X < 1) and P (X ≤ 1) are equal). We have 1 − F (1) = 1 − π2 tan−1 (1) =
1 − π2 · π4 = 1 − 12 = 12 .

(7) (original) Find the PDF f (x) for the CDF


0 √
 x ≤ −1
F (x) = 1 + 3 x −1 < x < 0

1 0 ≤ x,


and sketch pictures of the PDF and CDF.

4 JAMES ROWAN

We find the PDF by differentiating. For x ≤ −1, the PDF is f (x) = 0, and for 0 ≤ x, the PDF is also
f (x) = 0, since F is constant on those regions. For −1 < x < 0, we have f (x) = 31 x−2/3 by the power rule.
Picture of the PDF (bottom) and CDF (top) are sketched below:

(8) (original) Find the PDF f (x) for the CDF


0
 x < −1
3
F (x) = x − 14 x3 + 1
−1 ≤ x < 1
4 2
1 1 ≤ x,


and sketch pictures of the PDF and CDF.

We find the PDF by differentiating. For x < −1, the PDF is f (x) = 0, and for 1 ≤ x, the PDF is also
f (x) = 0, since F is constant on those regions. For −1 < x < 1, we have f (x) = 43 − 43 x2 by the power
rule. Picture of the PDF (bottom) and CDF (top) are sketched below:
 MATH 10B DISCUSSION SECTION PROBLEMS 3/21 – SOLUTIONS 5

(9) (original, challenge) We define a random variable X as follows. We flip a fair coin once.
• If the coin comes up heads, X is distributed according to a continuous uniform distribution on [−1, 1].
• If the coin comes up tails, X = 0.
What is the CDF of X?

The CDF is defined as the function F (x) such that F (x) = P (X ≤ x). For x ≤ −1, P (X ≤ x) = 0. For
−1 ≤ x < 0, P (X ≤ x) = 41 + x4 , as there is a 12 chance that X is drawn according to a uniform distribution
and the CDF for the uniform distribution on [−1, 1] is F (x) = 21 + x2 . At x = 0, P (X ≤ 0) = 34 , as there
is a 12 chance that x = 21 . For 0 ≤ x ≤ 1, P (X ≤ x) = 34 + x4 , as there is a 12 chance that X is drawn
according to a uniform distribution and the CDF for the uniform distribution on [−1, 1] is F (x) = 12 + x2 ,
plus a 21 chance that X = 0 since the coin came up tails. For x ≥ 1, F (x) = 1. Note that this CDF has a
jump at x = 0.
6 JAMES ROWAN

2. Acknowledgments
All problems labeled “Stewart/Day” taken from Day, Troy and Stewart, James. Biocalculus: Calculus, Proba-
bility, and Statistics for the Life Sciences. Cengage Learning, 2019.