CMI

Solutions to 2015 Entrance Examination for BSc Programmes at CMI
Part A Solutions
1. Ten people sitting around a circular table decide to donate some money for charity. You
are told that the amount donated by each person was the average of the money donated
by the two persons sitting adjacent to him/her. One person donated Rs. 500. Choose the
correct option for each of the following two questions. Write your answers as a sequence
of two letters (a/b/c/d).
What is the total amount donated by the 10 people?

(a) exactly Rs. 5000 (b) less than Rs. 5000 (c) more than Rs. 5000
(d) not possible to decide among the above three options.
What is the maximum amount donated by an individual?
(a) exactly Rs. 500 (b) less than Rs. 500 (c) more than Rs. 500
(d) not possible to decide among the above three options.
Answer: exactly Rs. 5000, exactly Rs. 500. Consider the person who donated Rs. 500.
Suppose the neighbor to the left donates 500 + x. Then the one on the right donates
500 − x. But continuing leftward, the amounts donated are 500 + 2x, 500 + 3x, . . ., forcing
x to be 0, since you come around to the neighbor to the right.
2. Consider all finite letter-strings formed by using only two letters A and B. We consider
the usual dictionary order on these strings. See below for the formal rule with examples.
Formal rule: To compare two strings w1 and w2 , read them from left to right. We say
“w1 is smaller than w2 ” or “w1 < w2 ” if the first letter in which w1 and w2 differ is A in
w1 and B in w2 (for example, ABAA < ABB by looking at the third letters) or if w2 is
obtained by appending some letters at the end of w1 (e.g. AB < ABAA).
For each of the statements below, state if it is true or false. Write your answers as a
sequence of three letters (T for True and F for False) in correct order.
(a) Let w be an arbitrary string. There exists a unique string y satisfying both the
following properties: (i) w < y and (ii) there is no string x with w < x < y .
Answer: True, append A to w.
(b) It is possible to give an infinite decreasing sequence of strings, i.e. a sequence
w1 , w2 , . . . , such that wi+1 < wi for each positive integer i.
Answer: True. B, AB, AAB, AAAB, . . .
(c) Fewer than 50 strings are smaller than ABBABABB.
Answer: False. There are infinitely many such strings e.g. A, AA, AAA, AAAA,
....
1
3. A positive integer n is called a magic number if it has the following property: if a and b
are two positive numbers that are not coprime to n then a + b is also not coprime to n.
For example, 2 is a magic number, because sum of any two even numbers is also even.
Which of the following are magic numbers? Write your answers as a sequence of four
letters (Y for Yes and N for No) in correct order.
(i) 129 (ii) 128 (iii) 127 (iv) 100.
Answer: Only 128 and 127 are magic numbers. See that n is a magic number if and
only if n is a power of a prime. (Otherwise, write n = ab with a, b coprime.)
4. Let A, B and C be unknown constants. Consider the function f (x) defined by
f (x) = Ax2 + Bx + C when x ≤ 0 ,

= ln(5x + 1) when x > 0 .
Write the values of the constants A, B and C such that f 00 (x), i.e., the double derivative
of f , exists for all real x. If this is not possible, write “not possible”. If some of the
constants cannot be uniquely determined, write “not unique” for each such constant.
Answer: The only problem is at x = 0. For continuity, ln(0 + 1) = C. For f 0 (0) to exist,
f must be continuous and the left and right derivatives of f at x = 0 (which are easily
seen to exist) must match, i.e. 5 = B. For f 00 (0) to exist, f 0 (0) must exist and left and
right derivatives of f 0 at x = 0 must match, i.e. 2A = −52 . So A = − 25 2
, B = 5, C = 0.
5. Consider the polynomial p(x) = (x + a1 )(x + a2 ) · · · (x + a10 ) where ai is a real number

for each i = 1, . . . , 10. Suppose all of the eleven coefficients of p(x) are positive. For each
of the following statements, decide if it is true or false. Write your answers as a sequence
of four letters (T/F) in correct order.
(i) The polynomial p(x) must have a global minimum. (ii) Each ai must be positive.
(iii) All real roots of p0 (x) must be negative. (iv) All roots of p0 (x) must be real.
Answer: All are true. (i) The degree is even, so p(x) goes to +∞ as x → ±∞. So p(x)
must attain a global minimum somewhere by continuity. (ii) The roots of p(x) are −ai .
By positivity of coefficients of p(x), no nonnegative number is a root of p(x). Thus all
−ai are negative, so all ai > 0. (iii + iv) All 10 roots of p(x) are real and negative. There
is a root of p0 (x) between consecutive roots of p(x) (this is valid even in case of multiple
roots). So all 9 roots of p0 (x) are real and negative as well. For negativity, one can also
note that all coefficients of p0 (x) are positive and apply the logic in (ii) to p0 (x).
6. Fill in the blanks. Let C1 be the circle with center (−8, 0) and radius 6. Let C2 be the
circle with center (8, 0) and radius 2. Given a point P outside both circles, let `i (P ) be
the length of a tangent segment from P to circle Ci . The locus of all points P such that
`1 (P ) = 3 `2 (P ) is a circle with radius and center at ( , ).
Answer: Center = (10,0), radius = 6. Use the distance formula and the Pythagorean
theorem to get y 2 +(x+8)2 −62 = 9(y 2 +(x−8)2 −4). Simplifying gives y 2 +(x−10)2 = 62 .
Another way, assuming the locus to be a circle: note that the ratio of the radii of C1 , C2
and that of the tangents is the same (namely 3). Now use similar triangles to see that
2
the desired circle intersects the X-axis at coordinates 4 and 16, giving a diameter of the
desired circle (why?)
√ 10 √
7. (i) By the binomial theorem ( 2+1)10 = Ci ( 2)i , where Ci are appropriate constants.
P
√ i=0
Write the value of i for which Ci ( 2)i is the largest among the 11 terms in this sum.
√
Ci+1 ( 2)i+1
Answer: i = 6. One way: simplify the ratio and see that this ratio is > 1 till
√
Ci ( 2)i
i = 5 and < 1 from i = 6 onwards.
√ √
(ii) For every natural number n, let ( 2+1)n = pn + 2 qn , where pn and qn are integers.
Calculate lim ( pqnn )10 .
n→∞
√ n
√
Answer: 32. Using binomial expansion see that √ ( 2 − 1) = ±(p n −
√ 2 qn ), where the
n
sign depends
√ on the parity of n.√ As n → ∞, ( 2 − 1) → 0 since ( 2 − 1) < 1. Thus
(pn − 2 qn ) → 0 and so pqnn → 2.
8. The format for car license plates in a small country is two digits followed by three vowels,
e.g. 04 IOU . A license plate is called “confusing” if the digit 0 (zero) and the vowel O are
both present on it. For example 04 IOU is confusing but 20 AEI is not. (i) How many
distinct number plates are possible in all? (ii) How many of these are not confusing?
Answer: (i) 102 × 53 = 12500. (ii) 102 × 43 plates without vowel O + 92 × (53 − 43 )
plates with vowel O but without 0. This gives 6400 + 4941 = 11341.
9. Recall that sin−1 is the inverse function of sin, as defined in the standard fashion. (Some-
times sin−1 is called arcsin.) Let f (x) = sin−1 (sin(πx)). Write the values of the following.
(Some answers may involve the irrational number π. Write such answers in terms of π.)
R 2.5
(i) f (2.7) (ii) f 0 (2.7) (iii) 0 f (x) dx
(iv) the smallest positive x at which f 0 (x) does not exist.
Answer: The graph of f is periodic with period 2. From x = −0.5 to x = 0.5 it is the
line y = πx of slope π passing through the origin and from x = 0.5 to x = 1.5 it is the
line with slope −π, which crosses the X axis at x = 1. Using this we see that (i) f (2.7) =
sin−1 (sin(2.7π)) = Rsin−1 (sin(0.7π))R = sin−1 (sin(0.5π + 0.2π)) = 0.5π − 0.2π = 0.3π. (ii)
2.5 2.5
f 0 (2.7) = −π (iii) 0 f (x) dx = 2 f (x) dx = π/8 and (iv) the smallest positive x at
which f 0 (x) does not exist is x = 1/2.
10. Answer the three questions below. To answer (i) and (ii), replace ? with exactly one of
the following four options: < , = , > , not enough information to compare.
(i) Suppose z1 , z2 are complex numbers. One of them is in the second quadrant and the
other is in the third quadrant. Then |z1 | − |z2 | ? |z1 + z2 |.
(ii) Complex numbers z1 , z2 and 0 form an equilateral triangle. Then |z12 + z22 | ? |z1 z2 |.
(iii) Let 1, z1 , z2 , z3 , z4 , z5 , z6 , z7 be the complex 8-th roots of unity. Find the value of
Π (1 − zi ), where the symbol Π denotes product.
i=1,...,7
3
Answer: (i) |z1 | − |z2 | < |z1 + z2 |. One way: using triangle inequality for z1 + z2
and −z2 we get |z1 | ≤ |z1 + z2 | + | − z2 | and so |z1 | − |z2 | ≤ |z1 + z2 |. Now we may
take absolute value on the LHS because switching z1 and z2 keeps RHS the same. For
equality, z1 + z2 and −z2 must point in the same direction, so z1 and z2 must be along
the same line. But they are in quadrants 2 and 3, so this cannot happen.
(ii) z2 must be obtained by rotating z1 by angle π/3, say in the counterclockwise direction
πi 2πi
(otherwise interchange the two). Then z2 = z1 e 3 . Then z12 + z22 = z12 (1 + e 3 ) and
πi 2πi πi
z1 z2 = z12 e 3 . Now 1 + e 3 = e 3 (see by calculation or picture), so we have in fact
z12 + z22 = z1 z2 .
x8 −1
(iii) We have Π (x−zi ) = x−1
= 1+x+. . .+x7 . Putting x = 1 gives Π (1−zi ) = 8.
i=1,...,7 i=1,...,7
11. There are four distinct balls labelled 1,2,3,4 and four distinct bins labelled A,B,C,D. The
balls are picked up in order and placed into one of the four bins at random. Let Ei denote
the event that the first i balls go into distinct bins. Calculate the following probabilities.
Notation: P r[X] = the probability of event X taking place. P r[X|Y ] = the probability
of event X taking place, given that event Y has taken place.
Answer:
4! 3 1 2! 1
(i) P r[E4 ] = 44
= 32
(ii)P r[E4 |E3 ] = 4
(iii) P r[E4 |E2 ] = 42
= 8
(iv) P r[E3 |E4 ] = 1.
Part B Solutions
1. Carefully solve the following series of questions. If you cannot solve an earlier part, you
may still assume the result in it to solve a later part.
(a) For any polynomial p(t), the limit lim p(t)
t is independent of the polynomial p. Justify
t→∞ e
this statement and find the value of the limit.
(b) Consider the function defined by
q(x) = e−1/x when x > 0 ,

= 0 when x = 0 ,
= e1/x when x < 0 .
Show that q 0 (0) exists and find its value. Why is it enough to calculate the relevant limit
from only one side?
(c) Now for any positive integer n, show that q (n) (0) exists and find its value. Here q(x)
is the function in part (b) and q (n) (0) denotes its n-th derivative at x = 0.
Answer: (a) If p(t) is constant, then the limit = 0. Otherwise we get a form ±∞
∞
. Using
p(t) p0 (t)
L’Hospital’s rule, we get lim et = lim et = 0 by induction on the degree of t (or apply
t→∞ t→∞
L’Hospital’s rule repeatedly).
4
−1/h
(b) The right side derivative = lim+ q(h)−q(0)
h
= lim+ e h
= lim+ e1/h
1/h = lim t
t. (Let
h→0 h→0 h→0 t→+∞ e
+
t = 1/h. As h → 0 , t → +∞.) This limit is 0, e.g. by part (a).
Now q is an even function, so letting k = −h, the left side derivative = lim− q(h)−q(0)
h
=
h→0
lim+ q(−k)
−k
= lim+ q(k)
−k
. Using the earlier calculation this also equals −0 = 0.
k→0 k→0
Note: It is wrong to argue that q 0 (0) = lim q 0 (x) because to do so, we first need to know
x→0
that q 0 is continuous at 0, but we have not even shown that q 0 (0) exists! For the same
reason it is wrong to argue below that q (n) (0) = lim q (n) (x).
x→0
(c) We will show by induction on n that q (n) (0) = 0. The case n = 1 is done. (We can
even start the induction at n = 0 by interpreting q (0) (x) = q(x).) Assuming that we are
(n) (n) (0)
done up to n and to prove the statement for n+1, we need to calculate lim q (h)−q
h
=
h→0
(n)
lim q h(h) , because q (n)
(0) = 0 by induction. Therefore it is good to examine q (n) (h) for
h→0
h 6= 0. This is easy to calculate by the usual rules, but the formulas will be different for
positive and negative h. For h 6= 0, as q is even, q 0 is odd, so q 00 is even, etc. and in
general q (n) (h) = (−1)n q (n) (−h). Therefore, just as for part (b), it suffices to show that
(n)
lim+ q h(h) = 0. By another induction, we see easily that for h > 0, q (n) (h) = p(1/h)e−1/h
h→0
for some polynomial p. Proof: q 0 (h) = ( h12 )e−1/h . Assuming the result for n, we have

q (n+1) (h) = [p(1/h)e−1/h ]0 = − h12 (−p(1/h) + p0 (1/h))e−1/h , which has the desired form.
(n) (h) −1/h
So we have lim+ q h
= lim+ p(1/h)e
h
= lim tp(t)e−t = lim tp(t)
et
= 0 by part (a). Here
h→0 h→0 t→∞ t→∞
we have again substituted t = 1/h.
2. Let p, q and r be real numbers with p2 + q 2 + r2 = 1.
(a) Prove the inequality 3 p2 q + 3 p2 r + 2 q 3 + 2 r3 ≤ 2.
(b) Also find the smallest possible value of 3 p2 q + 3 p2 r + 2 q 3 + 2 r3 . Specify exactly

when the smallest and the largest possible value is achieved.
Answer: We have 3 p2 q + 3 p2 r + 2 q 3 + 2 r3 = (q + r)(3p2 + 2q 2 + 2r2 − 2qr) =
(q +r)(3(p2 +q 2 +r2 )−(q 2 +r2 +2qr)) = (q +r)(3−(q +r)2 ) = x(3−x2 ) = 3x−x3 , where
x = q + r. Let us examine possible values of x in view of the constraint p2 + q 2 + r2 = 1.
We have 2qr ≤ q 2 + r2 e.g. because (q − r)2 ≥ 0. Adding q 2 + r2 , we get 2 2
√≤
√ q + r + 2qr
2 2 2 2 2 2 2 2
2q +2r ≤ 2, because q +r ≤ p +q +r = 1. Thus (q+r) ≤ 2. So − √ 2 ≤ q+r ≤ 2.
Note that equalities are achieved precisely when p = 0 and q = r = ±1/ 2.
Thus altogether
√ √ we have to find extrema of the odd function f (x) = 3x − x3 over the
0
interval [− 2, 2]. The √ critical points
√ are when f (x) = 3 − 3x2 = 0, i.e. x = ±1. Thus
we need to see only f (± 2) = ± 2 and f (±1) = ±2. Therefore −2 ≤ 3 p2 q + 3 p2 r +
2 q 3 + 2 r3 ≤ 2. Moreover, 3 p2 q + 3 p2 r + 2 q 3 + 2 r3 = ±2 precisely when x = q + r = ±1.
In each case, this gives a line segment in the qr-plane joining (±1, 0) and (0, ±1). Note
that both these segments lie within the circle q 2 + r2 = 1, so each point on them leads
to two valid points (p, q, r) on the unit sphere.
5
3. (a) Show that there are exactly 2 numbers a in {2, 3, . . . , 9999} for which a2 −a is divisible
by 10000. Find these values of a.
(b) Let n be a positive integer. For how many numbers a in {2, 3, . . . , n2 − 1} is a2 − a
divisible by n2 ? State your answer suitably in terms of n and justify.
Answer: (a) We have 10000 = 16 × 625 as product of prime powers. Recall the notation
a|b, meaning b is divisible by a. We have 10000|a2 − a if and only if (625|a(a − 1) and
16|a(a − 1)). Because a and a − 1 cannot share a factor, in turn this is equivalent to
having both the conditions (1) 625|a or 625|a − 1 AND (2) 16|a or 16|a − 1. Now if
the coprime integers 16 and 625 both divide the same natural number (in our case a or
a − 1), their product 10000 will also divide this number. In our case this would force
a = 0, 1, or ≥ 10000, all of which are not allowed. Thus the given requirement on a is
equivalent to having either (1) 16|a and 625|a − 1 OR (2) 16|a − 1 and 625|a. Each case
has a unique solution, respectively a = 9376 and a = 625 (e.g. use modular arithmetic:
in case 1, we have a = 625k + 1, which is k + 1 mod 16, forcing k = 15 because 16|a and
a ∈ {2, 3, . . . , 9999}).
(b) Let n = p1 e1 . . . pk ek be the factorization of n into powers of distinct primes. The
analysis in part (a) tells that required values of a are obtained as follows: write n2 = xy
as a product of two coprime integers and find values of a in {2, 3, . . . , n2 − 1} that are
simultaneously 0 mod x and 1 mod y. These are precisely the values of a that we want.
This is because each p2e i
i
must divide a or a − 1, as a and a − 1 are coprime.
Now the Chinese remainder theorem tells you that there is always an a that is 0 mod x
and 1 mod y. Moreover it is unique modulo xy = n2 because difference between any two
solutions would be divisible by xy.
The total number of ways to write n2 = xy as a product of coprime integers is exactly
2k as it amounts to choosing which of the k distinct primes to include in x and then
the rest go into y. (Notice that x and y are not interchangeable.) However, we have to
delete the two cases x = 1, y = n2 and y = 1, x = n2 , as these will respectively lead to
solutions a = 1 and a = 0 or n2 , which are not in {2, 3, . . . , n2 − 1}. Finally it is easy to
see that different choices of x lead to different values of a. This is because, of the primes
p1 , . . . , pk in the factorization of n, precisely the ones dividing x will divide a and the
remaining primes will not, because they divide a − 1.
Thus the final answer is 2k − 2. Note that this matches with the special case in part (a).
Finally, note that there was nothing special about taking a square: instead of n2 it could
be any positive integer m and we would proceed the same way to find requisite integers
a in {2, 3 . . . , m − 1} based on prime factorization of m.
4. Let f : R → R be a twice differentiable function, where R denotes the set of real numbers.
Suppose that for all real numbers x and y, the function f satisfies
f 0 (x) − f 0 (y) ≤ 3|x − y|.
Answer the following questions. No credit will be given without full justification.
(a) Show that for all x and y, we must have |f (x) − f (y) − f 0 (y)(x − y)| ≤ 1.5(x − y)2 .
6
(b) Find the largest and smallest possible values for f 00 (x) under the given conditions.
Answer: (a) Note that the given inequality stays valid if we take absolute value of the
LHS, because we may interchange x and y without affecting RHS.
Fix x, y and let t = x − y. For now let x ≥ y, i.e. t ≥ 0. For h ∈ [0, t], the value of y + h
0 0
varies between y and x. We are given that R t |f 0(y + h) − f 0(y)| ≤ 3|h|.
R t Integrate with
respect to h over the interval [0, t] to get 0 |f (y + h) − f (y)|dh ≤ 0 3|h|dh = 1.5t2 .
The LHS should remind us of the following general fact: the absolute value of a definite
integral ≤ theR tdefinite integral of absoluteR value the same function over the same interval.
t
So we get | 0 f (y + h) − f 0 (y)dh| ≤ 0 |f 0 (y + h) − f 0 (y)|dh. Combining with the
0
Rt
previous inequality we have | 0 f 0 (y + h) − f 0 (y)dh| ≤ 1.5t2 . Finally we calculate the
Rt Rt
LHS. | 0 f 0 (y + h)dh − 0 f 0 (y)dh| = |f (y + t) − f (y) − f 0 (y)t|, where the first integral
is calculated using the fundamental theorem of calculus and the second one is just the
integral of the constant f 0 (y). Substituting x − y for t gives the desired result.
Notes: (1) If x < y, then tR< 0. We use the same strategy but all definite integrals should
0
be taken over [t, 0]. Now t 3|h|dh = 1.5t2 = LHS of the desired inequality. In the final
R0 0
calculation we get t f (y + h) − f 0 (y)dh = f (y) − f (y + t) + f 0 (y)t = negative of the
previous answer. So when we take absolute value of this integral, we again get the same
RHS of the desired inequality. (2) f 0 (y + h) − f 0 (y) and its absolute value are integrable
functions of h because they are continuous. This is because f 0 (y + h) is just a shift of
the function f 0 and f 0 is continuous because it is differentiable by hypothesis.
0 0 0 0
(b) We have, for x 6= y, | f (x)−f
x−y
(y)
| ≤ 3, so −3 ≤ f (x)−f
x−y
(y)
≤ 3. Taking limit as y → x,
we get −3 ≤ f 00 (x) ≤ 3. It is easy to provide examples where f 00 attains the extreme
values ±3, e.g. f (x) = ±1.5x2 . These satisfy the hypothesis and have constant f 00 = ±3.
5. For an arbitrary integer n, let g(n) be the GCD of 2n + 9 and 6n2 + 11n − 2. What is
the largest positive integer that can be obtained as the value of g(n)? If g(n) can be
arbitrarily large, state so explicitly and prove it.
Answer: Long division gives 6n2 + 11n − 2 = (2n + 9)(3n − 8) + 70. By Euclidean
algorithm, GCD(6n2 + 11n − 2, 2n + 9) = GCD(2n + 9, 70). Thus g(n) divides 70. But
since g(n) divides 2n + 9, which is odd, g(n) divides 35. When n = 13, 2n + 9 = 35 and
hence g(13) = 35. Thus the maximum value of g(n) is 35. (Precisely for which n do we
have g(n) = 35 (or, if you wish, 1 or 5 or 7)? A bit more work will tell you. Try it.)
6. You are given the following: a circle, one of its diameters AB and a point X.
(a) Using only a straight-edge, show in the given figure how to draw a line perpendicular
to AB passing through X. No credit will be given without full justification. (Recall that
a straight-edge is a ruler without any markings. Given two points, a straight-edge can
be used to draw the line passing through the given points.)
Answer: Line AX cuts the circle in C. Line BX cuts the circle in D. Lines AD and BC
intersect in E. Line XE is perpendicular to line AB. Reason: Angles ADB and ACB are
right angles, being angles in a semicircle. The altitudes of triangle XAB are concurrent.
Two of them are AD and BC, so the third is contained in line XE. (Notice that we always
use lines rather than line segments - this is important for part (b).)
7
(b) Do NOT draw any of your work for this part in the given figure. Reconsider your
procedure to see if it can be made to work if the point X is in some other position, e.g.,
when it is inside the circle or to the “left/right” of the circle. Clearly specify all positions
of the point X for which your procedure in part (a), or a small extension/variation of it,
can be used to obtain the perpendicular to AB through X. Justify your answer.
Answer: Case 1: Suppose X is not on the line AB (so XAB is a triangle), nor on the
tangents to the circle at A (so line XA meets the circle in a point C different from A),
nor on the tangent at B (so line XB meets the circle in a point D different from B) nor on
the given circle (so C, D and X are all different). In this case the exact same procedure
will work so long as we understand that the altitudes and their intersection point may
lie outside triangle XAB. This is because the lines XA and XB meet the circle in two
distinct points C and D that are different from X,A and B.
Case 2: Suppose X is on one of the two tangents, say the tangent at A, but X is different
from A. In this case XA itself is the desired line! In terms of the construction, here we
have A = C = E. Of course we have to assume that we can detect whether a line meeting
a circle does so in one point or two. But this assumption is implicit in Case 1 also,
because there we need to be able to identify the second point of intersection!
Case 3: If X is on line AB, then XAB is not a triangle. If X is not on line AB but
X is on the circle, then XAB is a triangle but X = C = D = E, so we cannot draw
line XE. Thus in these cases, the above procedure fails. Nonetheless even in these
cases it is possible to draw a perpendicular through X to line AB using only
a straightedge. It is a challenge to you to find a suitable procedure!
8
Answers to Part A
1. If K comes second, then L was third (one correct answer for R). But then R would also
need to be second (one correct answer for M), a contradiction. So K cannot be second.
So M must have won, etc. The order is M R L K.
GDP
2. Per-capita GDP is population . Letting G and P denote the old GDP and population
1.078G
respectively, the new per-capita GDP is (1+x)P where x is the unknown percent change
in population we wish to calculate. The percent increase in per-capita GDP is 10% = 0.1.
So we have 1.078
1+x
= 1.1. Solving for x we get 1 + x = 1.078
1.1
98×11
= 100×11 = 0.98. So x is −0.02.
So population decreased by 2%.
3. Given n = pq = 110179. The number of integers relatively prime to n and smaller than
n is (p − 1)(q − 1). So we have pq − p − q + 1 = 109480. We get p + q = 700. Now p, q
are solutions to the quadratic
√ √ x2 − 700x + 110179. The discriminant of this quadratic is
490000 − 440716 = 49284 = 22. So we get p = 700+2222
= 461 and q = 700−222
2
= 239.
4. Let there be a steps to the right (east), b steps north-west and c steps southwest. The
total number of steps is a+b+c. The key idea is to think of the northwest step as a move
in the complex plane along ω, the complex cube root of unity, the southwest step as a
move in the complex plane along ω 2 and the step to the right as a move along ω 3 = 1.
From the hypothesis we then have a + bω + cω 2 = 1. Using 1 + ω + ω 2 = 0 we see that
a − 1 = b = c. This then rules out a + b + c = 6, so the number of 6 step paths is zero.
A 7 step path is possible only
with a = 3, b = 2, c = 2. The number of such paths is the
7
multinomial coefficient 3,2,2 = 210. (Instead of complex numbers one can also think in
terms of vector addition in the plane.)
5. Let θ = 101 π
. Let A = cos2 (θ) + cos2 (2θ) + · · · + cos2 (100θ). Let B = sin2 (θ) + sin2 (2θ) +
2
· · · + sin (100θ). We have A + B = 100, and A − B = cos(2θ) + cos(4θ) + · · · + cos(200θ).
π 2π 2πi
Since θ = 101 , we see that cos(2θ) = cos( 101 ) is the real part of e 101 , i being the complex
square root of −1. Interpreting the other terms in A − B similarly we see that A − B is
the real part of the sum of the 101th roots of unity except 1. So A − B = −1. This gives
A = 99 2
, B = 1012
.
6. The given function is defined using the two functions x2 + 1 and tan(x). Both these
functions are continuous wherever they are defined. Since every irrational number z has
a non terminating, non repeating decimal expansion we see that given any > 0 there is
a rational number p such that the distance between z and p is less than . Using these
facts one can see that the given function will be continuous precisely at those x in the
interval [0, 4π] where x2 + 1 = tan(x). Since x2 + 1 is positive, it will intersect tan(x)
exactly once in the intervals [0, π2 ], [π, 3π
2
], [2π, 5π
2
], [3π, 7π
2
], as tan(x) increases from 0 to
∞ in each of these intervals. tan(x) is negative elsewhere in the given domain. So we
have 4 points of continuity.
1
7. TTFF
Since the set S is nonempty, there is an element m ∈ S. But then m = m + 0 and
so 0 ∈ S. 1 cannot be in S, otherwise it will contain all non-negative integers. It is
not difficult to see by the division algorithm that if m, n are in S then so is their GCD.
Therefore two coprime numbers cannot be in S. Otherwise their GCD, which is 1, will be
in S, a contradiction. It follows that such sets S are precisely those of the form nZ≥0 , the
set of all non-negative multiples of a fixed non-negative integer n. So there are infinitely
many such possible sets.
8. TTFF
If g(x) is linear, it is 3x + 5 because the values at 1 and 2 are 8 and 11 respectively. If
g(x) is a polynomial then it is 3x + 5 plus a multiple of (x − 1)(x − 2) · · · (x − 15). So g(x)
cannot be a polynomial of degree 10. But it can be a polynomial of degree 15 or more.
g being differentiable does not mean that it is a polynomial. You can fit any number of
differentiable functions to the given data.
9. TFFT
i The mean value theorem tells us S ⊂ T .

ii T ⊂ S is false, example f (x) = sin(x). Here f 0 (0) = 1 is in T but not in S.
iii T = S = R can happen at points where f is not differentiable.
iv S has mean value property, because of continuity. (Why?)
10. TFFT
BP and CP are angle bisectors meeting at P , so AP bisects ∠A since the angle bisectors
are concurrent. The angles marked with symbol ◦ at point P are all 60◦ because ∠EP D =
twice this common value. It follows that half the sum of ∠B and ∠C is 60◦ . So ∠A is
60◦ . The others are false, in fact check that any triangle with ∠A = 60◦ , angle bisectors
BD and CE, their point of intersection P and P F bisecting ∠BP C will satisfy the given
data. All four statements are true if and only if the triangle ABC is equilateral.
Solutions to Part B.
1. Out of the 14 students taking a test, 5 are well prepared, 6 are adequately prepared and
3 are poorly prepared. There are 10 questions on the test paper. A well prepared student
can answer 9 questions correctly, an adequately prepared student can answer 6 questions
correctly and a poorly prepared student can answer only 3 questions correctly.
For each probability below, write your final answer as a rational number in lowest form.
(a) If a randomly chosen student is asked two distinct randomly chosen questions from
the test, what is the probability that the student will answer both questions correctly?
2
Note: The student and the questions are chosen independently of each other. “Random”
means that each individual student/each pair of questions is equally likely to be chosen.
(b) Now suppose that a student was chosen at random and asked two randomly chosen
questions from the exam, and moreover did answer both questions correctly. Find the
probability that the chosen student was well prepared.
Solution. (a) The probability that a randomly chosen student is well prepared is 5/14.
The probability of a well prepared student answering two randomly chosen questions
correctly is 92 / 10
2
. So the probability that a randomly chosen student is well prepared
5 (9)
AND answers two randomly chosen questions correctly is 14 × 102 = 27 . A student belongs
(2)
to exactly one of the three preparedness categories, so the desired probability is obtained
by adding 27 with the results of parallel calculations for the other two categories. We get
P (both answers correct) =

9 6 3

2 2 2
P (well prepared) 10
+ P (moderately prepared) 10
+ P (weakly prepared) 10 ,

2 2 2
which equals
5 36 6 15 3 3 31
× + × + × = .
14 45 14 45 14 45 70
(b) The probability that a randomly chosen student was well prepared given that he
answered both questions correctly is
P (well prepared and both correct) 2/7 20

P (well prepared|both correct) = = = .
P (both correct) 31/70 31
2. By definition the region inside the parabola y = x2 is the set of points (a, b) such that
b ≥ a2 . We are interested in those circles all of whose points are in this region. A bubble
at a point P on the graph of y = x2 is the largest such circle that contains P . (You may
assume the fact that there is a unique such circle at any given point on the parabola.)
(a) A bubble at some point on the parabola has radius 1. Find the center of this bubble.
(b) Find the radius of the smallest possible bubble at some point on the parabola. Justify.
Solution. A bubble at the point P = (a, a2 ) must be tangential to the parabola at

(a, a2 ). (Why?) It must also be symmetric with respect to Y-axis (why?) and so its
center O must be on the Y-axis. The radius OP of this bubble is perpendicular to the
common tangent to the parabola and to the bubble at P . The slope of this tangent =
2a, so the slope of radius OP = −1 2a
(for a 6= 0). Let Q = (0, a2 ). Using triangle OP Q,
slope of OP = −OQa
= −1
2a
. Therefore OQ = 12 , regardless of the value of a.
(a) By Pythagoras, OP 2 = ( 21 )2 + a2 = 1. So a2 = 3
4
and P = (0, 43 + 21 ) = (0, 54 ).
3
(b) For any nonzero a, the radius of the bubble satisfies OP 2 = ( 21 )2 + a2 , so OP > 12 .
The smallest bubble is at the origin and its radius is 12 . (One cannot just directly take
a = 0 in the above calculations. Argue by continuity or do a separate calculation at the
origin.)
3. Consider the function f (x) = xcos(x)+sin(x) defined for x ≥ 0.
(a) Prove that Z 1

0.4 ≤ f (x)dx ≤ 0.5.
0
√
Solution. It is easy to see that for 0 ≤ x ≤ 1, we have 1 ≤ cos(x) + sin(x) ≤ 2, and so
√
x1 ≥ xcos(x)+sin(x) ≥ x 2 .
As all three functions are non-negative in [0,1], we can integrate the inequalities over
that interval to get
Z 1
1 1 1
≥ f (x)dx ≥ √ > = 0.4.
2 0 2+1 1.5 + 1
(b) Suppose the graph of f (x) is being traced on a computer screen with the uniform
speed of 1 cm per second (i.e., this is how fast the length of the curve is increasing). Show
that at the moment the point corresponding to x = 1 is being drawn, the x coordinate
is increasing at the rate of
1
p cm per second.
2 + sin(2)
Rxp
Solution. Length of the curve from x = 0 to any given x is l(x) = 0 1 + f 0 (u)2 du.
dl
It is given that dt = 1 cm/second at all times. One needs to find dx
dt
when x = 1.
dl dl dx dl
p
By chain rule dt = dx dt
. By the fundamental theorem of calculus dx = 1 + f 0 (x)2 . We
calculate f 0 (1)
p = cos(1) + sin(1). (Use f√(x) = x
cos(x)+sin(x)
= eln x(cos(x)+sin(x)) , etc.) So at
dl
x = 1, dx = 1 + (cos(1) + sin(1))2 = 2 + sin 2. Chain rule gives the answer.
(Remark: We are using calculus to analyze what in reality is a discrete situation, as a

computer will draw pixel by pixel. So the whole description is an approximation. It is
also probably more realistic to assume dx
dt
to be constant.)
4. Let A be a non-empty finite sequence of n distinct integers a1 < a2 < · · · < an . Define
A + A = {ai + aj |1 ≤ i, j ≤ n},
4
i.e., the set of all pairwise sums of numbers from A. E.g., for A = {1, 4}, A+A = {2, 5, 8}.
(a) Show that |A + A| ≥ 2n − 1. Here |A + A| means the number of elements in A + A.

(b) Prove that |A+A| = 2n−1 if and only if the sequence A is an arithmetic progression.
(c) Find a sequence A of the form 0 < 1 < a3 < · · · < a10 such that |A + A| = 20.
Solution. (a) Easy induction, see answer to (b). Or explicitly, one has the 2n−1 distinct
numbers a1 + a1 < a1 + a2 < · · · < a1 + an < a2 + an < . . . < an + an in A + A. (A way
to visualize is to write ai + aj at point (i, j) in the XY-plane. Any step to the right or
up increases the number. To reach from 2a1 to 2an needs 2n − 1 such steps. The given
example is the path along bottom row and then rightmost column.)
(b) Suppose the ai form an arithmetic progression. Then for a fixed k, the value of
ai + ak−i is constant for all possible i, where 2 ≤ k ≤ 2n. For the converse use induction.
There is nothing to prove for n = 1, 2. For n > 2, remove an from A to get a set B. Now
|A + A| − |B + B| ≥ 2, because the two distinct numbers an−1 + an and 2an in A + A
are greater than all numbers in B + B. So for |A + A| = 2n − 1 to happen, one must
have |B + B| = 2n − 3, which by induction forces a1 , . . . , an−1 to be in an arithmetic
progression. Moreover an−2 + an must be in B + B and it can only be the largest number
2an−1 (because all others are smaller than an−2 + an ). This shows that an is the next
term of the same arithmetic progression.
(c) 0,1,2,3,4,5,6,7,8,10. This answer is unique. (Why?)
5. Find a polynomial p(x) that simultaneously has both the following properties.
(i) When p(x) is divided by x100 the remainder is the constant polynomial 1.
(ii) When p(x) is divided by (x − 2)3 the remainder is the constant polynomial 2.
Solution. Suppose a polynomial f (x) leaves a constant remainder r when divided by

the polynomial (x − c)k . Then f 0 (x) is divisible by (x − c)k−1 . The converse is also
true: suppose for a polynomial f (x), the derivative f 0 (x) is divisible by (x − c)k−1 , say
f 0 (x) = q(x)(x − c)k−1 . Then f (x) leaves a constant remainder when divided by (x − c)k .
One can see this e.g. by substituting u = (x − c) in q(x)(x − c)k−1 and integrating.
In the given problem p0 (x) must be divisible by x99 as well as by (x − 2)2 . Moreover any
polynomial whose derivative is divisible by x99 (x − 2)2 will leave constant remainders
when divided by either of x100 and (x − 2)3 . The simplest way to find one such p(x) is
to integrate Ax99 (x − 2)2 = A(x101 − 4x100 + 4x99 ) to get
!
x102 4x101 4x100
p(x) = A − + +B
102 101 100
and solve for constants A and B to ensure desired values

102 of the constant remainders. We
2 4×2101 4×2100
have p(0) = B = 1 and p(2) = A 102 − 101 + 100 + 1 = 2, which gives A.
5
Theoretical approach. Working through the following reasoning will be very useful
for your understanding of basic arithmetic/algebra. It explains how to implememt the
Chinese remainder theorem using the Euclidean algorithm for finding GCD. This theorem
states the following. One can always find an integer that leaves desired remainders when
divided by two coprime integers a and b.
Suppose we are required to find an integer that leaves remainder r when divided by
a and remainder s when divided by b. A way to achieve this systematically is to use
the Euclidean algorithm, which finds GCD of two numbers by repeated division with
remainder. This algorithm also enables one to write the GCD in the form xa + yb, where
the integers x, y can be found explicitly by backward substitution in the steps used to
calculate the GCD. If a and b are coprime, i.e. if their GCD is 1, then we can write
1 = xa + yb. This tells you that xa is 1 modulo b and yb is 1 modulo a. Therefore,
sxa + ryb is r modulo a and s modulo b.
The relevance for this problem is that the same reasoning applies for polynomials in one
variable, because in this setting too one has division with remainder. Because x100 and
(x − 2)3 do not share a common factor, you know without any work that a polynomial
with given properties must exist. The same algorithm as the previous paragraph (but
now with polynomials) gives a systematic way to find it. In the given problem we could
use a different trick because the specified remainders here were rather simple (constants).
But there is a conceptual way as well by implementing the Chinese remainder theorem.
6. Find all pairs (p, n) of positive integers where p is a prime number and p3 − p = n7 − n3 .
Solution. The given equation is p(p − 1)(p + 1) = n3 (n2 + 1)(n + 1)(n − 1). As the
factor p on the LHS is a prime, it must divide one of the factors n − 1, n, n + 1, n2 + 1 on
the RHS.
A key point to deduce is that p > n2 . One way to do this is as follows. The LHS
= p3 − p is an increasing function of p for p ≥ 1, e.g. because the derivative 3p2 − 1
is positive. So for any given n ≥ 1, there is exactly one real value of p for which
LHS = RHS. Trying p = n2 gives LHS = n6 − n2 < n7 − n3 = RHS, e.g. because
n7 − n3 − (n6 − n2 ) = (n6 − n2 )(n − 1) > 0.
As the prime p is greater than n2 , it cannot divide any of n − 1, n, n + 1. So p must divide

n2 + 1 and therefore we must have p = n2 + 1, again because p > n2 . Substituting this
in the given equation, we get (n2 + 1)n2 (n2 + 2) = n3 (n2 + 1)(n + 1)(n − 1). Canceling
common factors gives n2 + 2 = n3 − n, i.e. 2 = n3 − n2 − n. This has a unique integer
solution n = 2, e.g. because the factor n on the RHS must divide 2 and now one checks
that n = 2 works. So n = 2 and the prime p = n2 + 1 = 5 give a unique solution to the
given equation.
6
2017 Entrance Examination for the BSc Programmes at CMI
Solutions
Read the instructions on the front of the booklet carefully!
Part A. Write your final answers on page 3.
Part A is worth a total of (4 × 10 = 40) points. Points will be given based only on clearly
legible final answers filled in the correct place on page 3. Write all answers for a single question
on the designated line and in the order in which they are asked, separated by commas.
Unless specified otherwise, each answer is either a number (rational/ real/ complex) or, where
appropriate, one of the phrases “infinite”/“does not exist”/“not possible to decide”. Write in-
teger answers in the usual decimal form. Write non-integer rationals as ratios of two integers.
1. Consider the following construction in a circle. Choose points A, B, C on the given circle
such that ∠ABC is 60◦ and AB = BC. Draw another circle that is tangential to the
chords AB, BC and to the original circle.
Do the above construction in the unit circle to obtain a circle S1 . Repeat the process in
S1 to obtain another circle S2 . What is the radius of S2 ?
Solution. Consider the center O and diameter BD of the unit circle. It is easy to see
that S1 passes through D and its center E lies between O and D. Let r be the radius of
S1 , so length of ED is r. Consider the perpendicular from E to chord BA, meeting BA
in point F. Then length of EF is also r and therefore in the 30-60-90 triangle BEF, the
length of the hypotenuse BE is 2r. Thus 2 = BD = BE + ED = 3r, thus r = 32 . By
similarity, the radius of S2 is 23 × 23 = 49 .
2. 10 oranges are to be placed in 5 distinct boxes labeled U, V, W, X, Y. A box may contain
any number of oranges including no oranges or all the oranges. What is the number of
ways to distribute the oranges so that exactly two of the boxes contain exactly two
oranges each?
Solution. From the five distinct boxes, there are 10 ways to pick the two boxes that will
have 2 oranges each. We need to distribute the remaining 6 oranges in the remaining
three boxes such that none of the three boxes gets exactly 2 oranges. The possible
distributions are 6+0+0 (which can be done in 3 ways) or 5+1+0 (6 ways) or 4+1+1 (3
ways) or 3+3+0 (3 ways). Thus the required answer is 10 × (3 + 6 + 3 + 3) = 150.
√
3. Find the volume of the solid obtained when the region bounded by y = x, y = −x and
the line x = 4 is revolved around the x-axis. (It may be useful to draw the specified
region.)
√
Solution. From x = 0 to x = 1 we have x ≥ | − x|, so from x = 0 to x = 1 the volume
swept out by the part of the given region that lies below X-axis is included in the volume
1
swept out by the part above X-axis. So from x =√0 and x = 1 we just have to take the
volume obtained
√ by revolving the area below y = x. Similarly, from x = 1 to x = 4 we
have | − x| ≥ x, so here we just have to take the volume obtained by the revolving the
area below y = x. Thus the required volume is√obtained by adding volumes of two solids
of revolution around X-axis: area under y = x from x = 0 to x = 1 and area under
y = x from x = 1 to x = 4.
4. Positive integers a and b, possibly equal, are chosen randomly from among the divisors
of 400. The numbers a, b are chosen independently, each divisor being equally likely to
be chosen. Find the probability that gcd(a, b) = 1 and lcm(a, b) = 400.
Solution. 400 = 52 × 24 has (2 + 1) × (4 + 1) = 15 factors, so total number of pairs

(a, b) is 225. For a, b to be coprime, they should have no prime factor in common and
then their lcm is just their product, which is required to be 400. So there are only four
4
allowed pairs: (1,400), (400,1), (25,16) and (16,25). The probability is 225 .
5. Find all complex solutions to the equation:
x4 + x3 + 2x2 + x + 1 = 0.
Solution. It is easy to see that x4 + x3 + 2x2 + x + 1 = (x2 + 1)(x2 + x + 1).
6. Let g be a function such that all its derivatives exist. We say g has an inflection point
at x0 if the second derivative g 00 changes sign at x0 i.e., if g 00 (x0 − ) × g 00 (x0 + ) < 0 for
all small enough positive .
(a) If g 00 (x0 ) = 0 then g has an inflection point at x0 . True or False?

(b) If g has an inflection point at x0 then g 00 (x0 ) = 0. True or False?
(c) Find all values x0 at which x4 (x − 10) has an inflection point.
Solution. In (c), g 00 (x) = 20x3 − 120x2 = 20x2 (x − 6) and this changes sign only at
x = 6. Note that for this function, g 00 (0) = 0 but g 00 does not change sign at x = 0, thus
(a) is FALSE. On the other hand (b) is TRUE: Suppose for some g, the double derivative
g 00 changes sign at x0 . Then g 00 (x0 ) = 0 as g 00 is continuous (because g 00 is given to be
differentiable).
7. Write the values of the following.

Z 3
(a) |3x2 − 3| dx.
−3
Z t
0
(b) f (1) where f (t) = |3x2 − 3| dx.
0
Solution. (a) By symmetry we can calculate the definite integral from 0 to 3 and double
the answer. Note that |3x2 − 3| = 3x2 − 3 from x = 1 to 3 and |3x2 − 3| = 3 − 3x2 from
x = 0 to 1. So break the calculation at x = 1 etc.
2
(b) |3x2 − 3| is a continuous function so by the fundamental theorem of calculus, f 0 (1) =
|3 × 12 − 3| = 0
8. For this question write your answers as a series of four letters (Y for Yes and N for No)
in order. Is it possible to find a 2 × 2 matrix M for which the equation M~x = p~ has:
(a) no solutions for some but not all p~; exactly one solution for all other p~ ?
(b) exactly one solution for some but not all p~; more than one solution for all other p~ ?
(c) no solutions for some but not all p~; more than one solution for all other p~ ?
(d) no solutions for some p~, exactly one solution for some p~ and more than one solution
for some p~ ?
Solution. If M has nonzero determinant, then for any p~, we see that M~x = p~ has the
unique solution ~x = M −1 p~. If determinant of M is zero then we can make two cases. (i)
If M is the zero matrix, then M~x = p~ has infinitely many solutions for p~ = ~0 and no
solutions otherwise. (ii) If M is nonzero then it is easy to see that we are solving two
equations in two variables whose left hand sides are proportional. So if the two right
hand constants that make up p~ are in the same proportion, then we will have infinitely
many solutions (because one of the variables can be arbitrary). If the constants are not
in the same proportion, then the two equations will be inconsistent and we will have no
solutions. Thus the answer is NNYN. It is also possible to think geometrically in terms of
(at most) two lines, each moving in a parallel family. If the lines have the same slope they
either coincide or don’t intersect. Otherwise they have a unique point of intersection.
Note: In general linear algebra gives the right tools to analyze matrix equations, e.g. in
this problem we can say the following. If M = 0 then the space of solutions is either
empty or two-dimensional. If M 6= 0 then either there is a unique solution (precisely
when determinant 6= 0) or, when determinant is 0, the space of solutions is either empty
or one-dimensional. For larger matrices the possibilities are more complicated, but they
can be described precisely using the language provided by linear algebra.
9. Let f be a continuous function from R to R (where R is the set of all real numbers) that
satisfies the following property: For every natural number n
f (n) = the smallest prime factor of n.
For example, f (12) = 2, f (105) = 3. Calculate the following.
(a) limx→∞ f (x).

(b) The number of solutions to the equation f (x) = 2016.
Solution. f (x) will take value 2 for all even x. At the same time, primes provide an
increasing infinite sequence of positive integers for which f (x) = x. Thus limx→∞ f (x)
does not exist. By intermediate value theorem, for each prime p > 2016 there is an x
between p and p + 1 such that f (x) = 2016.
3
10. Consider the following function:
(
x2 cos( x1 ), x 6= 0,
f (x) =
a, x = 0.
(a) Find the value of a for which f is continuous.

Use this value of a to calculate the following.
(b) f 0 (0).
(c) lim f 0 (x).
x→0
Solution. cos( x1 ) is sandwiched between −1 and 1, so limx→0 f (x) = 0 = a makes

h2 cos( 1 )−0
f continuous. Now f 0 (0) = limh→0 h
h
= limh→0 h cos( h1 ) which is similarly 0.
Finally, for nonzero x, calculate f 0 (x) = 2x cos( x1 ) + sin( x1 ), so limx→0 f 0 (x) does not exist
as limx→0 2x cos( x1 ) = 0 and limx→0 sin( x1 ) does not exist.
4
2017 Entrance Examination for the BSc Programmes at CMI
Solutions to Part B
1. Answer the following questions
(a) Evaluate
x
lim+ (xx − xx ).
x→0
First consider the limit

lim xx = lim+ (elogx x )
x→0+ x→0
log x
= lim+ (e 1/x ). (1)
x→0
Now consider the following limit

log x 1/x
lim+ = lim
x→0 1/x x→0 −1/x2
=0 (2)
substituting the value 0 from (2) in equation (1) we get that the limit is 1.
Now,
x x
lim+ (xx − xx ) = lim+ xx − lim+ xx
x→0 x→0 x→0
limx→0+ xx
= lim+ x − lim+ xx
x→0 x→0
=0−1
= −1.
1
2π
(b) Let A = 9
, i.e., A = 40 degrees. Calculate the following
1 + cos A + cos 2A + cos 4A + cos 5A + cos 7A + cos 8A.
There are many ways to arrive at the answer 1. Here are two approaches.
Let S be the above sum. Then
S = 1 + cos A + cos 2A + cos 4A cos(2π − 4A) + cos(2π − 2A) + cos(2π − A)
= 1 + 2(cos A + cos 2A + cos 4A)

3A A A
= 1 + 2 2 cos( ) cos( ) + cos(π − )
2 2 2

π A A
= 1 + 2 2 cos( ) cos( ) − cos( )
3 2 2

1 A A
= 1 + 2 2 × cos( ) − cos( )
2 2 2
= 1.
Recall that cos nA is the real part of einA . Then

8
X 2
X
S= cos nA − cos(3nA)
n=0 n=1
8 2
!
in 2π
X X
= Re einA − e 3
n=0
2
n=1
= Re 0 − ω − ω
= 1.
Here ω is a complex cube root of unity.

x
(c) Find the number of solutions to ex = 2017 + 1.
First, note that x = 0 is clearly a solution. Let f (x) = ex − 2017
x
− 1. Then
x0 = − log 2017 is the only critical point of f (x). For all x < x0 we have
f 0 (x) < 0. Since f (x) → ∞ as x → −∞ there is only one solution in the
interval (−∞, x0 ). For all x > x0 we have f 0 (x) > 0 (i.e., ex > 2017
1
). Hence
there is only one solution in the interval (x0 , ∞). In total there are exactly
two solutions.
2
2. Let L be the line of intersection of the planes x + y = 0 and y + z = 0.
(a) Write the vector equation of L, i.e., find (a, b, c) and (p, q, r) such that
L = {(a, b, c) + λ(p, q, r) | λ is a real number.}
(b) Find the equation of a plane obtained by rotating x + y = 0 about L by 45◦ .
Clearly the line L passes through the origin. Moreover L is in the direction
perpendicular to the normals to the both the planes. The direction vector can
be obtained by computing following cross product
(î + ĵ) × (ĵ + k̂) = î − ĵ + k̂.
Hence L can be written as

L = {(0, 0, 0) + λ(1, −1, 1) | λ is a real number }
First, note that the equation of any plane that contains the line L is given by
x + (1 + λ)y + λz = 0.
Second, note that one can rotate the plane x + y = 0 in either clockwise or in
anticlockwise direction. Consequently there are two such planes. The normal of
one of the planes makes an angle of 45◦ with the normal of x + y = 0 and the
other normal makes an angle of 135◦ .
π
(î + ĵ) · (î + (1 + λ)ĵ + λk̂) = ±|î + ĵ||î + (1 + λ)ĵ + λk̂| cos( )
p 4
2 + λ = ± 1 + (1 + λ) + λ 2 2
2
λ − 2λ − 2 = 0
√
λ = 1 ± 3.
So the equation of the plane is

√
x + y + (1 ± 3)(y + z) = 0.
3
3. Let p(x) be a polynomial of degree strictly less than 100 and such that it does not have
x3 − x as a factor. If
d100

p(x) f (x)
100 3
=
dx x −x g(x)
for some polynomials f (x) and g(x) then find the smallest possible degree of f (x). Here
d100
dx100
means taking the 100th derivative.
Using the division algorithm we have
p(x) r(x)
3
= q(x) + 3 (3)
x −x x −x
As the degree of q(x) is strictly less than that of p(x) its 100th derivative is
certainly zero. As x3 − x is not a factor of p(x) one may assume (without loss of
generality) that x2 − 1 is a divides r(x). In that case we have
d100 d100 k

p(x)
= 100
dx100 x3 − x dx x
100! k
= 100
x
Hence the least possible degree of f (x) is 0.
If one assumes that x3 − x doesn't divide p(x) then we have
r(x) A0 B0 C0
= + + .
x3 − x x x−1 x+1
Consequently,
f (x) A B C
= 101 + 101
+
g(x) x (x − 1) (x + 1)101
∴ f (x) = A(x2 − 1)101 + B(x2 + x)101 + C(x2 − x)101

202 201 101 101
= (A + B + C)x + 101(B − C)x + ( B+ C − 101A)x200 + · · · .
2 2
Choosing B = C and A+B+C = 0 we see that the coecient of x200 is (101)(102) 6=

0. Hence the least possible degree of f (x) in this case is 200.
4
4. The domain of a function f is the set of natural numbers. The function is defined as
follows: √
f (n) = n + n
where bkc denotes the nearest integer smaller than or equal to k. For example, bπc =
3, b4c = 4. Prove that for every natural number m the following sequence contains at
least one perfect square
m, f (m), f 2 (m), f 3 (m), . . .
The notation f k denotes the function obtained by composing f with itself k times, e.g.,
f2 = f ◦ f.
If m is itself a square then we are done. So assume that m = k2 + j for 1 ≤ j ≤ 2k.
Hence we have f (m) = k2 + j + k. Consider the following two sets
A = {m a natural number | m = k2 + j and 0 ≤ j ≤ k}.
B = {m a natural number | m = k2 + j and k + 1 ≤ j ≤ 2k}.
Suppose m is in the set B . Then
f (m) = k 2 + j + k
= (k + 1)2 + (j − k − 1).
Hence f (m) is either a square or is in A. Thus it is enough to assume that m ∈ A.

In that case k2 < f (m) < (k + 1)2 , so bf (m)c = k. Therefore
f 2 (m) = (k + 1)2 + (j − 1).
This clearly implies that f 2j (m) = (k + j)2 .
5
5. Each integer is colored with exactly one of three possible colors - black, red or white -
satisfying the following two rules: the negative of a black number must be colored white,
and the sum of two white numbers (not necessarily distinct) must be colored black.
(a) Show that the negative of a white number must be colored black and the sum of
two black numbers must be colored white.
(b) Determine all possible colorings of the integers that satisfy these rules.
Suppose an integer n is colored white. Then (n + n) = 2n is black, so −2n is
white, so −2n + n = −n is black. Thus, the negative of a white number must be
colored black. Now suppose the integers n and m are both colored black. Then
−n and −m are both white, so −n − m is black, so n + m is white. Thus, the sum
of two black numbers must be colored white.
One possible coloring has all the integers colored red, since there are no condi-
tions on red numbers.
If that is not the case, let n be the smallest positive integer that is not colored
red. Suppose the number n is colored black. Then we claim the remaining
colors are all fully determined. Namely, the numbers of the form (3k + 1)n will
be black, the numbers of the form (3k + 2)n will be white, and the numbers of
the form (3k)n will be red, for all integers k. And all remaining colors will be
red. On the other hand, if the number n is colored white to begin with, then the
remaining numbers will be determined by the same rules, but with black and
white switched. Thus we have listed all possible colorings.
In order to prove the above claim, we rst prove one more rule the colors must
obey. Namely, that (*) The sum of a black number and a white number must
be colored red. Suppose n is black and m is white, and that n + m is black.
But then (n + m) + (−m) is the sum of two black numbers, and must be colored
white, which is a contradiction. Similarly, the sum of n and m cannot be white.
Therefore it must be red.
Using this rule, it is easy to see that the numbers of the form (3k + 1)n will be
black, the numbers of the form (3k + 2)n will be white, and the numbers of the
form (3k)n will be red, for all integers k. It remains to show that all numbers
that are not multiples of n are colored red.
We can prove this by contradiction. As before n is the smallest positive integer
that is not red, and it is colored black. Suppose m is the smallest positive integer
that is neither red nor a multiple of n. Then m = qn + r, where 0 < r < n is the
remainder when m is divided by n. We know this remainder is nonzero, since m
is not a multiple of n. We also know that q > 0, since m > n. Suppose m is white.
Then, because −n is white, we know m − n = (q − 1)n + r is black, which gives us
a smaller non-red positive integer that's not a multiple of n. On the other hand,
suppose m is colored black. Then −2n is black, so m − 2n = (q − 2)n + r is white.
If q > 1, this gives us a smaller positive non-red integer that's not a multiple of
n, which is a contradiction, provided q > 1. But if q = 1, and m − 2n = −n + r is
white, that means that n − r is black, another contradiction.
6
6. You are given a regular hexagon. We say that a square is inscribed in the hexagon if it
can be drawn in the interior such that all the four vertices lie on the perimeter of the
hexagon.
(a) A line segment has its endpoints on opposite edges of the hexagon. Show that it
passes through the center of the hexagon if and only if it divides the two edges in
the same ratio.
(b) Suppose a square ABCD is inscribed in the hexagon such that A and C are on the
opposite sides of the hexagon. Prove that center of the square is same as that of
the hexagon.
(c) Suppose the side of the hexagon is of length 1. Then find the length of the side of
the inscribed square whose one pair of opposite sides is parallel to a pair of opposite
sides of the hexagon.
P
D A
U Q
T R
C B
(d) Show that, up to rotation, there is a unique way of inscribing a square in a regular
hexagon.
(a) Suppose a segment AC meets with opposite sides P Q and T S of a hexagon

and O is the midpoint of AC . We show that
PA
AQ
= TC
CS
⇐⇒ O is the center of the hexagon.
If O is the center of the hexagon, consider triangles OAQ and OCS . By
the SAS test these are congruent. Similarly, triangles OAP and OCT are
congruent.
Conversely, suppose AQ
PA
= TCSC = k (say), then
P Q = T S =⇒ P A+AQ = T C +CS =⇒ AQ(k+1) = CS(k+1) =⇒ AQ = CS.
So 4AQO ∼ = 4CT O, so that OQ = OT . Also, ∠AOQ = ∠COT and ∠AOP =

∠COS , so Q, O and T are collinear.
(b) Next suppose we have inscribed a square ABCD in a hexagon P QRST U ,
with A on P Q, B on QR, C on ST and D on T U . We claim that 4AQB is
7
congruent to 4CT D. This will prove that both diagonals pass through the
center of the hexagon (using the criterion proved above).
Proof: We know that P A k ST and AC is a transversal. So ∠QAC = ∠T CA,
also ∠BAC = ∠DCA = 45◦ . So ∠QAB = ∠DCT .
Similarly, ∠QBA = ∠CDT . Also, ∠AQB = ∠CT D, since they are angles in
a regular hexagon. Moreover, AB = CD. As a result we get that 4QBA ∼ =
4T DC .
So we have QB = T D and QA = T C . This in turn implies that BR = DU
and P A = CS Thus,
QB TD PA SC
= and = .
BR DU AQ CT
Hence AC and BD pass through the center of the hexagon.
(c) Let DU = x so DP = 1 − x. Observe that DC = 2x sin 30 and DA = 2(1 −
x) sin 60. Since DC = DA we solving the equations for x we get x = √3+1 2
.
√ √
Consequently the side DC = 3( 3 − 1).
(d) Finally we want to show that there is a unique way of inscribing a square
in a regular hexagon.
Proof: It will be enough to show that the ratios QB BR
and QA
AP
are equal.
Suppose on the contrary that these ratios aren't equal.
Let ∠QAB = α and ∠QBA = β . Note that then ∠OAQ = 45◦ + α and
∠OBQ = 45◦ + β . Also, α + β = 60◦ , since ∠AQB = 120◦ .
Let A0 be a point on QR such that QA . Since 4BOA0 is isosceles,
0
A0 R
= QA
AP
∠OBA equals ∠OA B , so that
0 0
180◦ = ∠OBA0 +∠OBQ = ∠OBQ+∠OA0 B = ∠OBQ+∠OAQ = 45◦ +β +45◦ +α,
so α + β = 0◦ , a contradiction since α + β = 60◦ .
8
Solutions for the 2018 Entrance Examination for the BSc Programmes at CMI
Part A
1. Consider an equilateral triangle ABC with altitude 3 centimeters. A circle is inscribed

in this triangle, then another circle is drawn such that it is tangent to the inscribed
circle and the sides AB, AC. Infinitely many such circles are drawn; each tangent to the
previous circle and the sides AB, AC. The figure shows the construction after 2 steps.
B C
Find the sum of the areas of all these circles.

Answer: The radius of the ( rst) inscribed circle is 1. Its not hard to see that
that as you go on inscribing the circles the corresponding radii decrease by 1/3.
Let A denote the total area of these circles then
A = π(1)2 + π(1/3)2 + π(1/9)2 + · · ·
= π + π(1/3)2 [1 + (1/3)2 + (1/9)2 + · · · ]
= π + (1/9)A
= (9/8)π.
2. Consider the following function defined for all real numbers x

2018
f (x) = .
10 + ex
How many integers are there in the range of f ?
Answer: 201. Note that the for all values of x the function is strictly decreasing
and the graph lies above x axis. As x goes far left the denominator approaches
10 and the function value approaches 201.8. On the other hand, as goes far right
the denominator blows up and the function value approaches 0. Since this is a
continuous function by the intermediate value theorem all values in the interval
(0, 201.8) are assumed.
3. List every solution of the following equation. You need not simplify your answer(s).
√3
√
x + 4 − 3 x = 1.
√
Put t = 3 x, to get (t3 + 4) = (1 + t)3 . This leads to the quadratic t2 + t − 1 =√0.
Solve it and then take the cube root of the solutions. The answers are −2 ± 5.
1
4. Compute the following integral
Z π
2 dx
√ √
0 ( sin x + cos x)4
√
Pull cos2 x out from the denominator and then substitute u for tan x + 1. The
answer is 31 .
5. List in increasing order all positive integers n ≤ 40 such that n cannot be written in the
form a2 − b2 , where a and b are positive integers.
Answer: 1, 4 and all even numbers of the form 4k + 2
6. Consider the equation
z 2018 = 20182018 + i,
√
where i = −1.
(a) How many complex solutions does this equation have?

(b) How many solutions lie in the first quadrant?
(c) How many solutions lie in the second quadrant?
The equation has 2018 complex solutions. In the polar form the right hand side
of the equation can be expressed as reiθ , where θ is a very small positive angle.
Note that 2018 is 2 mod 4. Of the 2018 solutions of x2 018 = r, one each is on
positive and negative X -axis. The remaining 2016 are divided equally in the four
quadrants, 504 each. Now rotating these by the very tiny angle θ/2018 gives 505
each in the rst and third quadrant but still 504 in second and fourth.
7. Let x3 + ax2 + bx + 8 = 0 be a cubic equation with integer coefficients. Suppose both r
and −r are roots of this equation, where r > 0 is a real number. List all possible pairs
of values (a, b).
Plugging in r and −r in the equation we get r2 + b = 0 and ar2 + 8 = 0. Let the
third root be s, then expanding (x+r)(x−r)(x+s) and comparing it with the given
equation tells us that ab = 8. So the possible values of a, b are −1, −2, −4, −8, i.e.,
both a, b negative such that ab = 8.
8. How many non-congruent triangles are there with integer lengths a ≤ b ≤ c such that
a + b + c = 20?
It is clear that 1 < a ≤ b ≤ c < 10. Now, c < a + b and c = 20 − a − b implies
10 < a + b; this also means that b ≥ a and b ≥ 11 − a. Moreover, we also have
b ≤ 20 − a − b. One can further conclude that a ≤ 6, otherwise 7 ≤ b ≤ 6. So as a
ranges from 2 to 6 we have that b takes the following values a = 2, b = 9; a = 3, b =
8; a = 4, b ∈ {7, 8}; a = 5, b ∈ {6, 7}; a = 6, b ∈ {6, 7}. The total number of possible
triangles is 8.
9. Consider a sequence of polynomials with real coefficients defined by
p0 (x) = (x2 + 1)(x2 + 2) · · · (x2 + 1009)
2
with subsequent polynomials defined by pk+1 (x) := pk (x + 1) − pk (x) for k ≥ 0. Find the
least n such that
pn (1) = pn (2) = · · · = pn (5000).
Answer n = 2018. Note that deg p0 (x) = 2018 and deg pk (x) = 2018 − k . De ne
gn (x) = pn (x) − pn (1), hence gn (x) has degree 2018 − n and 5000 roots.
10. Recall that arcsin(t) (also known as sin−1 (t)) is a function with domain [−1, 1] and range
[− π2 , π2 ]. Consider the function f (x) := arcsin(sin(x)) and answer the following questions
as a series of four letters (T for True and F for False) in order.
(a) The function f (x) is well defined for all real numbers x.TRUE
(b) The function f (x) is continuous wherever it is defined. TRUE
(c) The function f (x) is differentiable wherever it is continuous. FALSE
This is a periodic function with period 2π. On [−π/2, π/2] it is identity and
on [π/2, 3π/2] it is negative identity. Hence the function is well-de ned and
continuous everywhere. However, it is not di erentiable at nonzero multiples of
π/2.
3
Answers to part B
If you need extra space for any problem,
continue on one of the colored blank pages at the end and write a note to that effect.
1. Answer the following questions-
(a) A natural number k is called stable if there exist k distinct natural numbers
a1 , . . . , ak , each ai > 1, such that
1 1
+ ··· + = 1.
a1 ak
Show that if k is stable then k + 1 is also stable. Using this or otherwise, find all
stable numbers. [5 marks]
It is clear that 1 and 2 are not stable. However,P3 is stable. Let k ≥ 3 be
stable hence there are a1 , . . . , ak all distinct and a1i = 1. This implies that
1 1
= 1. Hence all numbers except 2 are stable.
P
2
+ 2ai
(b) Let f be a differentiable function defined on a subset A of R. Define
f ∗ (y) := max {yx − f (x)} ,

x∈A
whenever the above maximum is finite. For the function f (x) = − ln(x), determine
the set of points for which f ∗ is defined and find an expression for f ∗ (y) involving
only y and constants. [5 marks]
First, note that the function f (x) is de ned only for the positive values of
x. Now if y ≥ 0 then the rst derivative of xy + ln(x) is y + x1 which is
strictly positive for x > 0. Hence xy + ln(x) is an increasing function and
consequently f ∗ (y) is not de ned.
Now if y < 0 then x = − y1 is the only critical point of xy + ln(x). Moreover,
either of the derivative test tells us that it is in fact the maxima. Hence,
the domain of f ∗ (y) is y < 0 and
−1
f ∗ (y) = ln( ) − 1.
y
4
2. Answer the following questions
(a) Find all real solutions of the equation [6 marks]

x2 + x − 6
x2 − 2x = 1.
Explain why your solutions

√ are the only solutions.
Answer x = −3, 1, 1 ± 2 are the only solutions. First, we want either
x2 + x − 6 = 0 or x2 − 2x = 1. However, when x = 2 the base as well as the
exponent are 0 giving us an indeterminate form. Hence x = 2 will not work.
Moreover, when x = −3 the base is positive. Second, observe that when
x = 1 we get (−1)−4 which equals 1.
(b) The following expression is a rational number. Find its value. [9 marks]
√ √
q q
3 3
6 3 + 10 − 6 3 − 10.
Answer : 2. Let the numbers be a, b respectively. Note a3 − b3 = 20 and

ab = 2. Putting it in (a − b)3 we get (a − b)3 = 20 − 6(a − b). This cubic has
one real solution a − b = 2 and two complex solutions.
5
3. Let f be a function on the nonnegative integers defined as follows
f (2n) = f (f (n)) and f (2n + 1) = f (2n) + 1.
(a) If f (0) = 0, find f (n) for every n. [2 marks]

(b) Show that f (0) cannot equal 1. [4 marks]
(c) For what nonnegative integers k (if any) can f (0) equal 2k ? [9 marks]
Answer
(a) Suppose f (0) = 0 then f (1) = 1 and f (2) = f (f (1)) = f (1) = 1. It implies that
f (3) = 1 + 1 = 2 and f (4) = f (1) = 1. The pattern continues and we get that
if 2k + 1 ≥ 3 then f (2k + 1) = 2. On the other hand if 2k ≥ 4 then f (2k) = 1.
(b) Suppose f (0) = 1. We have f (0) = f (2 · 0) = f (f (0)) = f (1). But we also have
f (1) = f (0) + 1, a contradiction.
(c) Suppose f (0) = 2k . Then, 2k = f (2·0) = f (f (0)) = f (2k ), and f (2k +1) = f (2k )+
1 = 2k + 1. Notice that f (1) = f (0) + 1 = 2k + 1, and f (2) = f (f (1)) = 2k + 1.
In this way, we see that for any n, f (2n ) = 2k + 1. This contradicts that fact
that f (2k ) = 2k .
6
4. Let ABC be an equilateral triangle with side length 2. Point A0 is chosen on side BC
such that the length of A0 B is k < 1. Likewise points B 0 and C 0 are chosen on sides CA
and AB with AC 0 = CB 0 = k. Line segments are drawn from points A0 , B 0 , C 0 to their
corresponding opposite vertices. The intersections of these line segments form a triangle,
labeled P QR in the interior.
C
B0
R
A0
P Q
A B
C0
4(1−k)
Show that the triangle P QR is an equilateral triangle with side length √
k2 −2k+4
.
Note that triangles ABA0 , CAC 0 and BCB 0 are congruent by the SAS test. Trian-
gles BA0 Q, CB 0 R and AC 0 P are also congruent. By using the property of opposite
angles we get that all the three angles of the triangle P QR are the same. Hence
it is an equilateral triangle.
Dropping the perpendicular bisector AO on the side BC we get the following:
AA02 = AO2 + A0 A2
√
= (1 − k)2 + ( 3)2
= k 2 − 2k + 4.
Observe that triangles ABA0 and BQA0 are similar by the AAA test: A0 QB and
A0 BA are 60 degrees and A0 BQ and A0 AB are corresponding angles. Therefore:
AB BA0 A0 A
= =
BQ QA0 A0 B
√
2 k k 2 − 2k + 4
= =
BQ QA0 k
2k
BQ = √
k 2 − 2k + 4
k2
QA0 = √ .
k 2 − 2k + 4
Now using AA0 = AP + P Q + QA0 we get
4(1 − k)
PQ = √ .
k 2 − 2k + 4
7
5. An alien language has n letters b1 , . . . , bn . For some k < n/2 assume that all words
formed by any of the k letters (written left to right) are meaningful. These words are
called k-words. Such a k-word is considered sacred if:
i) no letter appears twice and,

ii) if a letter bi appears in the word then the letters bi−1 and bi+1 do not appear. (Here
bn+1 = b1 and b0 = bn .)
For example, if n = 7 and k = 3 then b1 b3 b6 , b3 b1 b6 , b2 b4 b6 are sacred 3-words. On the

other hand b1 b7 b4 , b2 b2 b6 are not sacred. What is the total number of sacred k-words?
Use your formula to find the answer for n = 10 and k = 4.
We will count the sacred words starting with b1 . Since b1 is chosen bn and b2 are
out of the picture. In order to ll the remaining k −1 positions we have to choose
non-consecutive bi 's. Note that, specifying these bi 's is equivalent to specifying
the gaps between them. For example, let n = 7, k = 3 and we would like to
choose b1 , b3 , b6 . Then the triple (1, 2, 1) speci es that leave one alphabet after
b1 , drop two after b3 and drop one after b6 . Hence, in general let x1 , x2 , . . . , xk be
these gaps. It is clear that each of this gap is at least 1 and they add up to n − k.
So our counting problem is now - in how many di erent ways one can choose k
natural numbers, each of which is at least 1, that add up to n − k. The answer
is n−k−1 . In fact, this is equivalent to counting the number of ways one can

k−1
choose k − 1 `plus' signs from n − k − 1of them when n − k is written as a sum of
1's (n − k of them). However, note that we haven't assigned positions to these
letters yet. t This can be done in (k − 1)! ways. Hence the answer is

n−k−1
(k − 1)! .
k−1
In order to count the total number of sacred words we just need to multiply the
above number by n. The nal answer is

n−k−1 (n − k − 1)!
n(k − 1)! = n(k − 1)!
k−1 (n − 2k)!(k − 1)!
(n − k − 1)!
=n
(n − 2k)!
= n (n − k − 1)(n − k − 2) · · · (n − 2k + 1).
For n = 10 and k = 4 the answer is 600.
8
6. Imagine the unit square in the plane to be a carrom board. Assume the striker is just
a point, moving with no friction (so it goes forever), and that when it hits an edge, the
angle of reflection is equal to the angle of incidence, as in real life. When it hits another
edge it bounces again similarly and so on. If the striker ever hits a corner it falls into
the pocket and disappears. The trajectory of the striker is completely determined by its
−−−→
starting point (x, y) and its initial velocity (p, q).
If the striker eventually returns to its initial state (i.e. initial position and initial velocity),
we define its bounce number to be the number of edges it hits before returning to its initial
state for the first time.
−−−→
For example, the trajectory with initial state [(.5, .5); (1, 0)] has bounce number 2 and
it returns to its initial state for the first time in 2 time units. And the trajectory with
−−−→
initial state [(.25, .75); (1, 1)] has bounce number 4.
−−−→
(a) Suppose the striker has initial state [(.5, .5); (p, q)]. If p > q ≥ 0 then what is the
velocity after it hits an edge for the first time? What if q > p ≥ 0? [2 marks]
(b) Draw a trajectory with bounce number 5 or justify why it is impossible. [3 marks]
−−−→
(c) Consider the trajectory with initial state [(x, y); (p, 0)] where p is a positive integer.
In how much time will the striker first return to its initial state? [2 marks]
−−−→
(d) What is the bounce number for the initial state [(x, y); (p, q)] where p, q are relatively
prime positive integers, assuming the striker never hits a corner? [8 marks]
(a) If p > q then the striker will hit the vertical edge rst, and its new velocity
−−−−→
will be (−p, q). If p < q then the striker will hit the horizontal edge rst, and
−−−−→
its new velocity will be (p, −q).
(b) No, it is not possible. If the striker has bounce number 5, then it must have
an odd number of vertical edge bounces or horizontal edge bounces. In the
former case, when the striker returns to its initial state, the x-component
of its velocity will be wrong, by the formula in part (a). In the latter case
the y component will be wrong.
(c) It will take p2 time to return to its initial state.
(d) The bounce number is 2p + 2q. At time 2, the striker will have completed p
horizontal round-trips and q vertical round trips, and will have returned to
its initial state. To see this, note that from part (c) it will take time p2 for
each horizontal round trip and time 2q for each vertical round trip. Since p
and q are relatively prime, it will only be at time 2 that an integer number
of vertical and horizontal round trips have been completed.
9
2019 Entrance Examination for the national undergraduate programmes at CMI
Solutions to Part A
1. For a natural number m, define Φ1 (m) to be the number of divisors of m and for k ≥ 2
define Φk (m) := Φ1 (Φk−1 (m)). For example, Φ2 (12) = Φ1 (6) = 4. Find the minimum k
such that
Φk (20192019 ) = 2.
Answer: 6
2. Let f be a real valued continuous function defined on R satisfying
f 0 (tan2 θ) = cos 2θ + tan θ sin 2θ, for all real numbers θ.

π
If f ( 0) = − cos 12 then find f (1).
Note there was a typo in the exam; it was printed f 0 (0) instead of f (0). An-
swer: Put y = tan2 θ. Then we have
f 0 (y) = 1
Hence the answer is 1 − cos 12π .

3. You have a piece of land close to a river, running straight. You are required to cut off a
rectangular portion of the land, with the river forming one of the sides of the rectangle so,
your fence will have three sides to it. You only have 60 meters of fencing. The maximum
area that you can enclose is . . . . . . .
Answer: 450 square meters.
4. The sum
S = 1 + 111 + 11111 + · · · + 11 · · · 1}
| {z
2k + 1
is equal to . . . . . . .
102k+3 − 99k − 109
Answer: .
99 × 9
5. You are given an 8×8 chessboard. If two distinct squares are chosen uniformly at random
find the probability that two rooks placed on these squares attack each other. Recall that
a rook can move either horizontally or vertically, in a straight line.
2
Answer: .
9
6. For how many natural numbers n is n6 + n4 + 1 a square of a natural number?
Answer: 1 (n = 2 is the only solution.
1
7. A broken calculator has all its 10 digit keys and two operation keys intact. Let us
call these operation keys A and B. When the calculator displays a number n pressing
A changes the display to n + 1. When the calculator displays a number n pressing B
changes the display to 2n. For example, if the number 3 is displayed then the key strokes
ABBA changes the display in the following steps 3 → 4 → 8 → 16 → 17.
If 1 is on the display what is the least number of key strokes needed to get 260 on the
display?
Answer: 9, there are exactly two sequences, for example, BBBBBBABB .
8. Let π = π1 π2 . . . . . . πn be a permutation of the numbers 1, 2, 3, . . . , n. We say π has its

first ascent at position k < n if π1 > π2 . . . > πk and πk < πk+1 . If π1 > π2 > . . . >
πn−1 > πn we say π has its first ascent in position n. For example when n = 4 the
permutation 2134 of has its first ascent at position 2.
The numberof permutations which have their first ascent at position k is . . . . . . .
Answer: nk (n − k)! − k+1
n
(n − k − 1)!.
For questions 9 and 10 below, some statements are given. For each statement,
state if it is true or false. Write your answer to each question as a sequence
of three/ four letters (T for True and F for False) in correct order.
9. Consider f : R × R → R defined as follows:

1
f (a, b) := lim loge [ena + enb ].
n→∞ n
(a) f is not onto i.e. the range of f is not all of R.
(b) For every a the function x 7→ f (a, x) is continuous everywhere.
(c) For every b the function x 7→ f (x, b) is differentiable everywhere.
(d) We have f (0, x) = x for all x ≥ 0.
Answer: FTFT
10. Let f : R → R.
R1
(a) There is no continuous function f for which f (x)(1 − f (x))dx < 41 .
0
R1
(b) There is only one continuous function f for which 0 f (x)(1 − f (x))dx = 14 .
R1
(c) There are infinitely many continuous functions f for which 0 f (x)(1−f (x))dx = 41 .
Answer: FFT
Solutions to Part B
1. For a natural number n denote by Map(n) the set of all functions f : {1, 2, . . . , n} →
{1, 2, . . . , n}. For f, g ∈ Map(n), f ◦ g denotes the function in Map(n) that sends x to
f (g(x)). [10 marks]
2
(a) Let f ∈ Map(n). If for all x ∈ {1, . . . , n} f (x) 6= x, show that f ◦ f 6= f .
Answer: Suppose f (f (x)) = f (x). Set y = f (x). Then we have f (y) = y , a
contraction.
(b) Count the number of functions f ∈ Map(n) such that f ◦ f = f .
Answer: Note that from the above part it follows that each x has to map
to a xed point of f (i.e., a y such that f (y) = y. So in order to count the
number of such functions we rst need to decide the number of xed points.
The number functions that have exactly k xed points is nk kn−k . In order
to get the total number sum the previous quantity over 1 ≤ k ≤ n.
2019
p w√of the equation z
2. (a) Count the number of roots − 1 = 0 over complex numbers
that satisfy |w + 1| ≥ 2 + 2. [5 marks]
Answer: Such roots can be expressed as follows
cos(2πk) sin(2πk)
w= +i for k = 0, ±1, . . . , ±1009.
2019 2019
Therefore,
cos(2πk)
|w + 1|2 = 2 + 2 .
2019
Hence we want to identify k such that
cos(2πk) 1
≥√ .
2019 2
Which is equivalent to
2πk π
| |≤
2019 4
i.e., |k| ≤ 252.
So there are 505 solutions satisfying the given inequality.
(b) Find all real numbers x that satisfy following equation: [5 marks]
8x + 27x 7
x x
= .
12 + 18 6
Answer: Put a = 2x and b = 3x . This reduces the given equation to the
following quadratic -
6a2 − 13ab + 6b2 = 0.
Solving the above equation and re-substituting we get x = ±1.
R∞
3. Evaluate 0 (1 + x2 )−(m+1) dx, where m is a natural number. [10 marks]
Answer: There are various ways to solve this. One can start with the subsitution
x = tan u, that changes the integral to
1 2π
Z
I= (cos(u))2m du.
4 0
3
Using integration by reduction technique the nal answer is
2π (2m)!
· 2m .
4 2 (m!)2
4. Let ABCD be a parallelogram. Let O be a point in its interior such that ∠AOB +
∠DOC = 180◦ . Show that ∠ODC = ∠OBC. [10 marks]
Answer: Note that there exists an external point P such that AP is parallel to
DO, BP is parallel to CO and OP is parallel to BC . Now AOBP is a cyclic
quadrilateral. Rest is a straightforward calculation involving angles.
5. Three positive real numbers x, y, z satisfy
x2 + y 2 = 32
y 2 + yz + z 2 = 42
√
x2 + 3xz + z 2 = 52 .
√
Find the value of 2xy + xz + 3yz. [10 marks]
Answer: Consider the right angled triangle ABC with sides 3, 4, 5 and an interior
point O such that AO = x, ∠AOB = 90 and CO = z, ∠COA = 150 and BO =
y, ∠BOC = 120. Then the three given equations are in fact cosine rule for each
of the triangle prescribed above. For example, in ∆BOC we have
42 = y 2 + z 2 − 2yz cos(120)
= y 2 + z 2 + yz.
The area of δABC (which is 6) calculated using the sine formula (for each of the
smaller triangle) gives us
1 1 1
6 = xy + yz sin 60 + sin 30
2 2 2
So the answer is 24.
Z ex
d 4
6. (a) Compute log(t) cos (t) dt . [4 marks]
dx 0
Rx
(b) For x > 0 define F (x) = 1 t log(t)dt. [6 marks]
Answer: This is fairly straightforward: substitute ey for t and use the Leibniz
rule for the di erntial under an integral sign to get the answer
(ex )(log(ex )) cos4 (ex ).
i. Determine the open interval(s) (if any) where F (x) is decreasing and the open
interval(s) (if any) where F (x) is increasing.
ii. Determine all the local minima of F (x) (if any) and the local maxima of F (x)
(if any) .
4
Answer:
Z x
0 d
F (x) = t log t dt
dx 1
= x log x.
Therefore F 0 (1) = 0. Moreover, F 00 (x) = 1 + log x. Hence one concludes that

F is decreasing on (0, 1), increasing on (1, ∞) and has a local minima at
x = 1.
5
Part A solutions
Part A is worth a total of 40 (= 4 × 10) points. Unless specified otherwise, each answer
is either a number (rational/real/complex) or, where appropriate, ∞ or −∞. If a desired
answer “does not exist” or is “not possible to decide”, state so. Write integer answers in the
usual decimal form. Write non-integer rationals as ratios of two integers.
A1. Each student in a small school has to be a member of at least one of THREE school
clubs. It is known that each club has 35 members. It is not known how many students are
members of two of the three clubs, but it is known that exactly 10 students are members of
all three clubs. What is the largest possible total number of students in the school? What
is the smallest possible total number of students in the school?
Answer: Let a, b, c be the numbers of students that are members of exactly one of the clubs
and x, y, z the numbers of students with double membership. Labelling a suitable Venn
diagram appropriately, we have a + x + y = b + y + z = c + x + z = 25. We want to find
max/min values of n = a + b + c + x + y + z + 10. Adding the three constraints we get
a + b + c + 2x + 2y + 2z = 75, giving n = 85 − (x + y + z). The maximum possible n
cannot be more than 85 and it is achieved when x = y = z = 0 and a = b = c = 25. For
minimum n we need to maximize x + y + z, which can be at most 37.5 (if we could take
a = b = c = 0), hence at most 37 as it needs to be an integer. This is achieved by, say
a = 1, b = c = 0, x = y = 12, z = 13 giving minimum n = 38 + 10 = 48.
A2. Let P be the plane containing the vectors (6, 6, 9) and (7, 8, 10). Find a unit vector that
is perpendicular to (2, −3, 4) and that lies in the plane P. (Note: all vectors are considered
as line segments starting at the origin (0, 0, 0). In particular the origin lies in the plane P.)
Answer: The desired vector is of the form v = t(6, 6, 9) + s(7, 8, 10) = (6t + 7s, 6t + 8s, 9t +
10s) and we need 0 = v · (2, −3, 4) = 2(6t + 7s) − 3(6t + 8s) + 4(9t + 10s) = 30t + 30s = 0.
Taking s = 1, t = −1, we get (1, 2, 1) and scaling it to a unit vector gives v = ±( √16 , √26 , √16 ).
A3. Calculate the following two definite integrals. It may be useful to first sketch the graph.
e2 1
ln |x|
Z Z
ln |x| dx dx
1 −1 |x|
R
Answer: Integrating by parts, ln |x| dx = x ln |x|−x+C, hence by the second fundamental
R e2 2
theorem of calculus, 1 ln |x| dx = (x ln |x| − x) |e1 = e2 + 1. For the second integral,
because of the discontinuity at 0, we need to break up the calculation there and take a limit.
2
Antiderivative of lnx|x| is (ln |x|)
2
. As x → 0+ , ln x → −∞. So the integral from 0 to 1 is −∞
and it is the same on the negative side because the integrand is even. So the final answer is
−∞ (which is better than saying undefined/does not exist).
1
A4. A fair die is thrown 100 times in succession. Find probabilities of the following events.
(i) 4 is the outcome of one or more of the first three throws.
(ii) Exactly 2 of the last 4 throws give an outcome divisible by 3 (i.e., outcome 3 or 6).
91
Answer: (i) 1 − ( 56 )3 = 216 (ii) 42 ( 13 )2 ( 23 )2 = 27
8

. .
A5. Write your answers to each question below as a series of three letters Y (for Yes) or N
(for No). Leave space between the group of three letters answering (i), the answers to (ii)
and the answers to (iii). Consider the graphs of functions
x3 x2 − x x3 − x
f (x) = 2 g(x) = h(x) = 3 .
x −x x3 x +x
(i) Does f have a horizontal asymptote? A vertical asymptote? A removable discontinuity?
(ii) Does g have a horizontal asymptote? A vertical asymptote? A removable discontinuity?
(ii) Does h have a horizontal asymptote? A vertical asymptote? A removable discontinuity?
Answer: (i) NYY (ii) YYN (iii) YNY
A rational function is continuous wherever it is defined and its only discontinuities are where
the denominator vanishes (so f is discontinuous at x = 0 and x = 1 and g and h only
at x = 0). Such a discontinuity is removable precisely when the rational function has a
(finite) limit at that x-value; otherwise the rational function has a vertical asymptote there,
3
because in that case the limit on either side must be ±∞. Now f (x) = x2x−x has a removable
discontinuity at (0, 0), vertical asymptote at x = 1 and f approaches ±∞ as x → ±∞, so
f has no horizontal asymptote. g has a vertical asymptote at x = 0, so the discontinuity
at x = 0 cannot be removed. g(x) approaches 0 as x → ±∞, so g(x) has X-axis as the
3 −x
horizontal asymptote. h(x) = xx3 +x has a removable discontinuity at (0, −1), no vertical
asymptote and h approaches 1 as x → ±∞, so h has y = 1 as the horizontal asymptote.
A6. Recall the function arctan(x), also denoted as tan−1 (x). Complete the sentence:
arctan(20202019) + arctan(20202021) 2 arctan(20202020),
because in the relevant region, the graph of y = arctan(x) .

Fill in the first blank with one of the following: is less than / is equal to / is greater than. Fill
in the second blank with a single correct reason consisting of one of the following phrases:
is bounded / is continuous / has positive first derivative / has negative first derivative / has
positive second derivative / has negative second derivative / has an inflection point.
Answer: “is less than” because in the relevant region (here the interval (0, ∞)), the graph
“has negative second derivative”, which ensures that any chord stays below the graph. The
point on the graph of y = arctan(x) at x = 20202020 is above the midpoint of the chord
joining the points at x = 20202019 and x = 20202021.
2
A7. The polynomial p(x) = 10x400 + ax399 + bx398 + 3x + 15, where a, b are real constants,
is given to be divisible by x2 − 1.
(i) If you can, find the values of a and b. Write your answers as a = ,b = . If it
is not possible to decide, state so.
(ii) If you can, find the sum of reciprocals of all 400 (complex) roots of p(x). Write your
answer as sum = . If it is not possible to decide, state so.
Answer: (i) By factor theorem, as x−1 and x+1 both divide p(x), we must have respectively
p(1) = 10+a+b+3+15 = 0 and p(−1) = 10−a+b−3+15 = 0. This gives a = −3, b = −25.
(ii) If roots are ri , then
P
X1 product of all roots except ri −coefficient of x/leading coefficient 1
= i = =− .
ri product of all roots constant term/leading coefficient 5
A8. For a positive integer n, let D(n) = number of positive integer divisors of n. For
example, D(6) = 4 because 6 has four divisors, namely 1, 2, 3 and 6. Find the number of
n ≤ 60 such that D(n) = 6.
Answer: There are 9 such n because n must have prime factorisation of type p2 q (giving
D(n) = (2 + 1)(1 + 1) = 6), or of type p5 (giving D(n) = 5 + 1 = 6). There are 8 possibilities
of the first type: 22 ×(3, 5, 7, 11 or 13), 32 ×(2 or 5), 52 × 2. Only 25 is of the second type.
A9. Notice that the quadratic polynomial p(x) = 1 + x + 21 x(x − 1) satisfies p(j) = 2j for
j = 0, 1 and 2. A polynomial q(x) of degree 7 satisfies q(j) = 2j for j = 0, 1, 2, 3, 4, 5, 6, 7.
Find the value of q(10).
Answer: Looking at the hint, recall that ni=0 ni = 2n . So, taking xi := x(x−1)···(x−i+1)
P
i!
,
P7 x
the polynomial p(x) = i=0 i satisfies the given requirements (note that
for10an integer
x 10
value of x that is < i, the value of i is 0). Therefore p(10) = 2 − 8 − 9 − 10
10

10
=
1024 − 45 − 10 − 1 = 968. (Given n + 1 distinct numbers xi , and a list of numbers yi , there
always exists a polynomial p(x) of degree at most n satisfying p(xi ) = yi . One can write such
p(x) explicitly by Lagrange interpolation but in this particular case a trick could be used.
Such a polynomial p(x) is also unique because the difference between two such polynomials
would have degree at most n but the difference would also have all n + 1 numbers xi as its
roots, forcing the difference to be the zero polynomial.)
A10. Note that 25 × 16 − 19 × 21 = 1. Using this or otherwise, find positive integers a, b

and c, all ≤ 475 = 25 × 19, such that
• a is 1 mod 19 and 0 mod 25,
• b is 0 mod 19 and 1 mod 25, and
• c is 4 mod 19 and 10 mod 25.
3
(Recall the mod notation: since 13 divided by 5 gives remainder 3, we say 13 is 3 mod 5.)
Answer: a = 400 = 25 × 16 = 1 + 19 × 21 is 1 mod 19 and 0 mod 25. Next, −399 =

−19 × 21 = −25 × 16 + 1 is 0 mod 19 and 1 mod 25, so take b = −399 + 475 = 76. Finally
4 × 400 + 10 × 76 = 2360 is 4 mod 19 and 10 mod 25, so take c = 2360 − 4 × 475 = 460. It is
easy to check that these work. This is an example of Chinese remainder theorem in action.
One can use the Euclidean algorithm to express gcd(25, 19) = 1 in the form 25 × 16 − 19 × 21,
but here this information is already given to you, so the answers can be just read off.
Part B solutions
Clearly explain your entire reasoning. No credit will be given without reasoning. Partial
solutions may get partial credit.
B1. [7 points] Suppose A, B, C, D are points on a circle such that AC and BD are diameters
of that circle. Suppose AB = 12 and BC = 5. Let P be a point on the arc of the circle from
A to B (the arc that does not contain points C and D). Let the distances of P from A, B, C
and D be a, b, c and d respectively. Find the values of a+b
c+d
and a−b
d−c
. You may assume d 6= c
so the second ratio makes sense.
Answer: Since AC and BD are diameters, all angles of ABCD are√ right angles (being angles
in a semicircle) and so ABCD is a rectangle. Its diameter is 13 = 122 + 52 . Use Ptolemy’s
theorem in PADB to get 12d = 13a + 5b and in PACB to get 12c = 5a + 13b. Add the two
equations to get a+b
c+d
= 32 . Subtract to get a−b
d−c
= 32 .
2πi
B2. [7 points] Let z = e( n ) . Here n ≥ 2 is a positive integer, i2 = −1 and the real number
2π
n
can also be considered as an angle in radians.
(i) Show that n−1 k

(ii) Show that 8k=0 cos(40k + 1)◦ = 0, i.e.,
P P
k=0 z = 0.
cos(1◦ ) + cos(41◦ ) + cos(81◦ ) + cos(121◦ ) + · · · + cos(241◦ ) + cos(281◦ ) + cos(321◦ ) = 0.

1−z n 1−e2πi
Answer: (i) z = cos( 2π
n
) + i sin( 2π
n
) 6= 1 so (1 + z + · · · + z n−1 ) = 1−z
= 1−z
= 0.
2πi
(ii) Take the equation in part (i) for n = 9. Multiply it by e( 360 ) and take the real part. Or
use cos(A + B) = cos A cos B − sin A sin B on the LHS with B = 1◦ , take out cos(1◦ ) as a
common factor from half the terms and sin(1◦ ) from the rest, and show using (i) that the
cos part and sin part in parentheses are each 0.
B3. [10 points] A spider starts at the origin and runs in the first quadrant along the graph
of y = x3 at the constant speed of 10 unit/second. The speed is measured along the length
of the curve y = x3 . Thepformula for the curve length along the graph of y = f (x) from
Rb
x = a to x = b is ` = a 1 + f 0 (x)2 dx. As the spider runs, it spins out a thread that is
always maintained in a straight line connecting the spider with the origin. What is the rate
in unit/second at which the thread is elongating when the spider is at ( 21 , 81 )?
4
You should use the following names for variables. At any given time t, the spider is at the
point (u, u3 ), the length of the thread joining it to the origin in a straight line is s and the
curve length along y = x3 from the origin till (u, u3 ) is `. You are asked to find ds dt
when
u = 21 . (Do not try to evaluate the integral for `: it is unnecessary and any attempt to do so
will not get any credit because a closed formula in terms of basic functions does not exist.)
Answer: s2 = u2 + u6√. Differentiate with respect to t to get 2s ds dt
= (2u + 6u5 ) du
dt
. Plug
17
1
in u = 2 to get s = 8 and dt = 4√17 dt at the point of interest. To calculate du
ds 19 du
dt
, we
d` d` du d`
use `. By chain rule dt = du dt . We are given dt = 10. By the fundamental theorem of
d`
√
calculus, du = 1 + 9u4 = 45 at u = 12 . Using these values first get du
dt
= 8 and then ds
dt
= √3817
unit/second.
B4. [12 points] Throughout this problem we are interested in real valued functions f satis-
fying two conditions: at each x in its domain, f is continuous and f (x2 ) = f (x)2 . Prove the
following independent statements about such functions. The hints below may be useful.
(i) There is a unique such function f with domain [0,1] and f (0) 6= 0.
(ii) If the domain of such f is (0, ∞), then (f (x) = 0 for every x) OR (f (x) 6= 0 for every x).
R∞
(iii) There are infinitely many such f with domain (0, ∞) such that 0 f (x) dx < 1.
Hints: (1) Suppose a number a and a sequence xn are in the domain of a continuous function
f and xn converges to a. Then f (xn ) must converge to f (a). For example f (0.5n ) → f (0)
1
and f (2 n ) → f (1) if all the mentioned points are in the domain of f . In parts (i) and (ii)
suitable sequences may be useful. (2) Notice that f (x) = xr satisfies f (x2 ) = f (x)2 .
Answer: Throughout the question the domain contains no negative real numbers, so all
numbers in the domain are squares and hence by f (x2 ) = f (x)2 , the output values of f can
only be nonnegative. (i) f (0)2 = f (0), so f (0) = 0 or 1, but we are given than f (0) 6= 0,
n
so f (0) = 1. Now take a ∈ (0, 1). Squaring repeatedly, we get the sequence a2 → 0. By
n n n
continuity of f , we have f (a2 ) → f (0) = 1. Since f (x2 ) = f (x)2 , we have f (a2 ) = f (a)2 ,
n
so f (a)2 → 1. For this to happen, the only possible (necessarily nonnegative) value of f (a)
is 1. Continuity forces f (1) = 1 as well. So only the constant function f (x) = 1 is possible.
√ √ p
(ii) Letting u = x2 in f (x2 ) = f (x)2 , we get f (u) = f ( u)2 , i.e., f ( u) = f (u). Suppose
f (a) = 0 and f (b) > 0 for positive real numbers a, b. Taking repeated square roots, we
−n −n −n −n
get two sequences a2 and b2 , both converging to 1. But f (a2 ) = f (a)2 → 0 and
−n −n
f (b2 ) = f (b)2 → 1, giving two values for f (1), which is a contradiction.
(iii) We may assume f (x) is never 0 on (0, ∞) and hence f (1) = 1. Since all powers of x
work we can splice suitable powers at x = 1 to get total integral less than 1. For finiteness,
in (0, 1] we must use xr with r > −1 and in [1, ∞), use xs with s < −1. This gives total
1 1
integral r+1 − s+1 , which is less than 1 for infinitely many combinations of r and s, e.g. take
r ≥ 1 and s < −3.
5
B5. [12 points] Consider polynomials p(x) with the following property, called (†).
(†) If r is a root of p(x), then r2 − 4 is also a root of p(x).
(i) We want to find every quadratic polynomial of the form p(x) = x2 + bx + c such that p(x)
has two distinct roots, has integer coefficients and has property (†). Prove that there are
exactly two such polynomials and list them in the provided space on a later page.
(ii) It is also true that there are exactly two cubic polynomials of the form p(x) = x3 +
ax2 + bx + c with the property (†) such that p(x) shares no root with the polynomials you
found in part (i). Explain fully how you will prove this along with the method to find the
polynomials, but do not try to explicitly find the polynomials.
Answer: (i) Let the set of two distinct roots be A = {r, s} and consider the function
f (x) = x2 − 4. The condition (†) ensures that this function maps the set A to itself. When
considered as a function on the set of roots, we will symbolically denote it by r → r2 − 4.
There are four possibilities for this function from A to A: (1) r → r and s → s (2) r → s
and s → r (3) r → s and s → s (4) r → r and s → r. In the first case, both r and s must be
roots of the polynomial x2 − x − 4, which does have two real solutions. In the second case we
have r → s → r, so f (f (r)) = (r2 − 4)2 − 4 = r and also f (f (s)) = s, so both r and s must
be roots of (x2 − 4)2 − 4 − x. This quartic polynomial MUST have x2 − x − 4 as a factor
(why?) and the roots of the remaining factor must automatically satisfy r → s → r (why?).
By long division one gets the other factor as x2 + x − 3, which also has two real roots.
In cases 3 and 4, one of the roots t is still a root of x2 − x − 4 and the other must and can be
−t, because f (x) = f (−x). However the resulting two polynomials of form (x − t)(x + t) do
not have integer coefficients, as can be seen explicitly by writing down the roots of x2 − x − 4.
(Note: The integer coefficients condition was mistakenly omitted in the exam, leading to
four polynomials instead of the claimed two. Therefore, part (i) was graded leniently and
anyone writing at least two correct polynomials, even in rough work, was given full credit.)
(ii) Similar, but now one has to argue that the only possibility is r → s → t → r with
r, s, t distinct. Proof: f (any root) must be a different root, because otherwise, this root
would satisfy the polynomial x2 − x − 4, contrary to the requirement of sharing no roots with
polynomials in part (i). Similarly for any two distinct roots, it is not possible for them to be
exchanged by f because then they would be roots of x2 + x − 3 from part (i). Now starting
with any root r, one must have f (r) = s distinct from r, then f (s) = t distinct from r and
s, and finally f (t) = r (because f cannot fix t, nor can it exchange t and s). So r, s, t satisfy
f (f (f (x))) = x, i.e., they are roots of ((x2 − 4)2 − 4)2 − 4 − x. Again this polynomial MUST
have x2 − x − 4 as a factor and dividing by it leaves a degree six polynomial, which MUST
factor into two cubics satisfying r → s → t → r. Justify these statements for yourself.
(Note: The missing condition in part (i) has no bearing on the answer to part (ii), but part
(ii) was still removed from the exam because its statement referred to the answer in part (i).
The few students who did valid work in part (ii) were unaffected by this decision.)
More complete picture for part (ii): In general (i.e. for condition (†) but now involving
6
a quadratic polynomial other than x2 − 4) the two cubics may have non-real roots but in
the given situation one can check that both cubics have real, in fact integer coefficients, e.g.
use an online tool such as www.wolframalpha.com/calculators/factoring-calculator to get
((x2 − 4)2 − 4)2 − 4 − x = (x2 − x − 4)(x3 − x2 − 6x + 7)(x3 + 2x2 − 3x − 5).
Since each cubic has at least one real root (by intermediate value theorem), all roots must be
real because the r → r2 − 4 condition cycles through all three roots. For a given polynomial
with, say, integer coefficients, which permutations of its roots are “valid” and which are not?
Such questions are properly formulated and studied in the famous Galois theory.
B6. [12 points] For sets S and T , a relation from S to T is just a subset R of S × T . If
(x, y) is in R, we say that x is related to y. Answer the following.
(i) A relation R from S to S is called antisymmetric if it satisfies the following condition: if
(a, b) is in R, then (b, a) must NOT be in R. For S = {1, 2, . . . , k}, how many antisymmetric
relations are there from S to S? (Parts (ii) and (iii) are independent of this part.)
Answer: (i) For antisymmetry, the k elements (i, i) of S × S are ruled out. The remaining
k(k − 1) elements divide into k2 pairs of the form (i, j) and (j, i). For each such pair, one
k
can choose (i, j) OR (j, i) OR neither to include in the relation. So the answer is 3(2) .
(ii) Write a recurrence equation for f (k, n) = the number of non-crossing relations from
{1, 2, . . . k} to {1, 2, . . . n} that have no isolated elements in either set. (See below for the
definitions of the two underlined terms and their visual meaning. Drawing pictures may be
useful.) Your recurrence should have only a fixed number of terms on the RHS.
(iii) Using your recurrence in (ii) or otherwise, find a formula for f (3, n).
Definition 1: We say that a relation from S to T has no isolated elements if each s in S is
related to some t in T and if for each t in T , some s in S is related to t.
Definition 2: We say that a relation R from {1, 2, . . . k} to {1, 2, . . . n} is non-crossing if
the following never happens: (i, x) and (j, y) are both in R with i < j but x > y.
Visual meaning: one can visualise a relation R very similarly to a function. List 1 to k as
dots arranged vertically in increasing order on the left and similarly list 1 to n on the right.
For each (s, t) in R, draw a straight line segment from s on the left to t on the right. In
the situation one wants to avoid for non-crossing relations, the segments connecting i with x
and j with y would cross. Having no isolated elements also has an obvious visual meaning.
Answer: (ii) The element 1 in each set has to be connected with 1 in the other. Now exactly
one of three things happens: 1 in S is connected to 2 in T , 2 in S is connected to 1 in T or 1
in neither set is connected to anything other than 1 in the other set. We get the recurrence
f (k, n) = f (k, n − 1) + f (k − 1, n) + f (k − 1, n − 1) with f (k, 1) = f (1, n) = 1. Draw a
picture to see how the non-crossing condition makes this work.
(iii) By the recurrence, get f (2, n) = 2n − 1 and then f (3, n) = 2n2 − 2n + 1, both for n ≥ 1.
7
2021 CMI BSc entrance examination
Part A Answers with explanation
1. Consider the two equations numbered [1] and [2]:
log2021 a = 2022 − a [1]

2021b = 2022 − b [2]
(a) Equation [1] has a unique solution.

(b) Equation [2] has a unique solution.
(c) There exists a solution a for [1] and a solution b for [2] such that a = b.
(d) There exists a solution a for [1] and a solution b for [2] such that a+b is an integer.
Correct options: a,b,d Wrong: c
a = 2021, b = 1 is the unique solution. (This can also be solved qualitatively, e.g., the
graph of y = 2022 − x is decreasing, with range the set of real numbers. It is easy
to see that it must intersect each of the graphs y = 2021x and y = log2021 x (both
of which are increasing) exactly once. Substituting c = 2022 − a, the first equation
is equivalent to 2021c = 2022 − c, which is the same as the second equation, so the
(unique) solutions c and b are equal, i.e., b = c = 2022 − a, so a + b = 2022. If a = b,
both would need to be 1011, which is manifestly not a solution to either equation.)
2. A prime p is an integer ≥ 2 whose only positive integer factors are 1 and p.
(a) For any prime p the number p2 − p is always divisible by 3.

(b) For any prime p > 3 exactly one of the numbers p − 1 and p + 1 is divisible by 6.
(c) For any prime p > 3 the number p2 − 1 is divisible by 24.
(d) For any prime p > 3 one of the three numbers p + 1, p + 3 and p + 5 is divisible
by 8.
Correct options: b,c Wrong: a,d
(a) is false for any number that is 2 modulo 3, in particular for 2. To see that (b) and
(c) are true, note that any prime p greater than 3 is not divisible by 3, so p is either
1 mod 3 (which makes p − 1 divisible by 3) or 2 mod 3 (which makes p + 1 divisible
by 3). Note also that both p − 1 and p + 1 are even, which gives (b). In fact they
are consecutive even numbers, so one of them is a multiple of 4, making their product
p2 −1 a multiple of 8, giving (c). Finally, note that any prime p that is 1 mod 8 violates
(d), e.g., p = 17.
1
3. We want to construct a triangle ABC such that angle A is 20.21◦ , side AB has length
1 and side BC has length x where x is a positive real number. Let N (x) = the number
of pairwise noncongruent triangles with the required properties.
(a) There exists a value of x such that N (x) = 0.

(b) There exists a value of x such that N (x) = 1.
(c) There exists a value of x such that N (x) = 2.
(d) There exists a value of x such that N (x) = 3.
Correct options: a,b,c Wrong: d
Draw a ray with endpoint A. The point C will be chosen on this ray later on. Draw a
segment AB of length 1 making an angle 20.21◦ with this ray. Now, to fulfil the required
properties, a necessary and sufficient condition for C is that it is on the original ray as
well as on the circle with center B and radius x. As x increases from 0, the number
of intersections of the expanding circle with the ray goes from 0 to 1 (when ABC is a
right angled triangle, i.e., when x = sin 20.21◦ ) to 2 and finally back to 1.
4. Consider polynomials of the form f (x) = x3 + ax2 + bx + c where a, b, c are integers.

Name the three (possibly non-real) roots of f (x) to be p, q, r.
(a) If f (1) = 2021, then f (x) = (x − 1)(x2 + sx + t) + 2021 where s, t must be integers.
(b) There is such a polynomial f (x) with c = 2021 and p = 2.
(c) There is such a polynomial f (x) with r = 21 .
(d) The value of p2 + q 2 + r2 does not depend on the value of c.
Correct options: a,d Wrong: b,c
(a) is true by the remainder theorem. Long division automatically gives integers s, t.
Uniqueness of quotient and remainder for polynomial long division means those are the
only values of s, t that work. (b) is false by substituting x = 2 into f (x) and noting that
c = 2021 forces f (2) to be odd, in particular nonzero. To see that (c) is false, substitute
x = 21 into f (x), multiply by 8 to clear denominators and see that the leading term
makes the integer 8f ( 12 ) odd. So f ( 21 ) is nonzero. (General version of (b) and (c) that
one gets by the similar reasoning: suppose a polynomial p(x) with integer coefficients
has a rational root rs written in lowest form. Then the leading coefficient of p(x) is
divisible by s and the constant term of p(x) is divisible by r. Often used special case: for
a polynomial p(x) = xn + lower terms with integer coefficients, any rational root must
be an integer.) For (d) note that p2 +q 2 +r2 = (p+q +r)2 −2(pq +pr +qr) = (−a)2 −2b
does not depend on c.
2
5. For any complex number z define P (z) = the cardinality of {z k |k is a positive integer},
i.e., the number of distinct positive integer powers of z. It may be useful to remember
that π is an irrational number.
(a) For each positive integer n there is a complex number z such that P (z) = n.
(b) There is a unique complex number z such that P (z) = 3.
(c) If |z| =
6 1, then P (z) is infinite.
(d) P (ei ) is infinite.
Correct options: a,c,d Wrong: b

2πi
(a) is true by z = e n or any primitive nth root of unity. (b) is false: there are 2
2πi
primitive third roots of 1, namely ω = e 3 and ω 2 . (c) is true because then each |z i |
is a distinct positive real number. (d) is true because P (z) is finite only if powers of
z cycle back to 1, which happens for z = reiθ only if (r = 1 and) the argument θ is a
rational multiple of π. But for z = ei , we have θ = 1.
6. A stationary point of a function f is a real number r such that f 0 (r) = 0. A polynomial

need not have a stationary point (e.g. x3 + x has none). Consider a polynomial p(x).
(a) If p(x) is of degree 2022, then p(x) must have at least one stationary point.
(b) If the number of distinct real roots of p(x) is 2021, then p(x) must have at least
2020 stationary points.
(c) If the number of distinct real roots of p(x) is 2021, then p(x) can have at most
2020 stationary points.
(d) If r is a stationary point of p(x) AND p00 (r) = 0, then the point (r, p(r)) is neither
a local maximum nor a local minimum point on the graph of p(x).
Correct options: a,b Wrong: c,d
(a) p0 (x) is a polynomial of degree 2021, which is odd, so it has a root by intermediate
value theorem by looking at behaviour as x → ±∞. (b) The graph of p(x) has to turn
between any two consecutive zeros, giving a stationary point, in fact a local max/min
(c) The graph of p(x) can turn more than once between zeros, or turn outside extreme
zeros or have stationary points that are not maxima or minima. (d) is false, e.g.,
p(x) = x4 .
3
7. Given three distinct positive constants a, b, c we want to solve the simultaneous equa-
tions
√
ax + by = 2
√
bx + cy = 3
(a) There exists a combination of values for a, b, c such that the above system has
infinitely many solutions (x, y).
(b) There exists a combination of values for a, b, c such that the above system has
exactly one solution (x, y).
(c) Suppose that for a combination of values for a, b, c, the above system has NO
solution. Then 2b < a + c.
(d) Suppose 2b < a + c. Then the above system has NO solution.
Correct options: a,b,c Wrong: d
Each of the given equations defines a line in the XY plane. (a) One can arrange
both lines to√ be identical by having each equation a scalar multiple of the other, e.g.,
a = 1, b = √32 , c = 32 . (b) There is a unique solution when the two lines are distinct
and not parallel. (c) The two lines are given to be parallel. So slopes are equal, i.e.,
b2 = ac. Thus b is the geometric mean of a and c, so b < the arithmetic mean a+c 2
.
2
(Recall that a, b, c are distinct and positive.) (d) is absurd. Just ensure b 6= ac.
8. Given two distinct nonzero vectors v1 and v2 in 3 dimensions, define a sequence of

vectors by
vn+2 = vn × vn+1 (so v3 = v1 × v2 , v4 = v2 × v3 and so on).
Let S = {vn |n = 1, 2, . . .} and U = { |vvnn | |n = 1, 2, . . .}. (Note: Here × denotes the

cross product of vectors and |v| denotes the magnitude of the vector v. The vector 0
with 0 magnitude, if it occurs in S, is counted. But in that case of course the 0 vector
is not considered while listing elements of U .)
(a) There exist vectors v1 and v2 for which the cardinality of S is 2.

(b) There exist vectors v1 and v2 for which the cardinality of S is 3.
(c) There exist vectors v1 and v2 for which the cardinality of S is 4.
(d) Suppose that for some v1 and v2 , the set S is infinite. Then the set U is also
infinite.
Correct options: b,c Wrong: a,d
4
It is easiest to do this geometrically, remembering that the cross product p×q of vectors
p and q is perpendicular to both of them and |p × q| = |p| |q| sin(angle between p
and q) = |p| |q| if p and q are perpendicular. The cross product of nonzero vectors is
zero if and only if the vectors are collinear. It is easy to see that the only way the zero
vector is in S is if v3 is zero, which will happen only when the nonzero vectors v1 and
v2 are collinear, and in that case the sequence is zero all the way from v3 onwards.
As the starting vectors v1 and v2 are distinct and nonzero, the third vector v3 = v1 ×v2 ,
being perpendicular to both v1 and v2 , is distinct from them. This is true even if v3
is 0 due to v1 and v2 being collinear. So (a) is false.
Basic calculation. Taking v1 = i and v2 = j, the sequence cycles: i, j, k, i, j, k, . . ., so

(b) is true. (c) is also true because we can arrange the sequence to be the following:
a vector not in {i, j, k}, j, k, i, j, k, . . . (e.g., take v1 = i + j and v2 = j. The basic
calculation repeats after v3 .) Note that any vector in the sequence depends only on
the previous two vectors.
(d) is false. The set S can easily be infinite (e.g., if you start with i and 2j, the mag-
nitudes of subsequent vectors will keep increasing), but U is always finite. First note
that the cycle of three vectors occurs whenever one starts with any two perpendicular
vectors of unit length. Now U consists of unit vectors in the direction of each nonzero
vector in S. So depending on the angle θ between v1 and v2 , the cardinality of U is
either 1 (when θ = 0), 2 (when θ = π), 3 (when θ = π/2) or 4 (in all other cases,
because v2 and v3 are still perpendicular).
9.
x x 4 + x6
f (x) = and g(x) = .
x + sin x e x − 1 − x2
(a) Limit as x → 0 of f (x) is 21 .
(b) Limit as x → ∞ of f (x) does not exist.
(c) Limit as x → ∞ of g(x) is finite.
(d) Limit as x → 0 of g(x) is 720.
Correct options: a,c Wrong: b,d
Calculate (a) and (c) using L’Hôpital’s rule. (Or in (a) use that sin x behaves like x
near 0 and in (c) the limit is 0 because ex dominates any polynomial for large x.) In
x
(b) the limit is 1 as f (x) is sandwiched between x±1 , both of which → 1. L’Hôpital’s
rule is not applicable as the expression one gets after attempting it does not have a
limit as x → ∞, so L’Hôpital’s rule does not tell us anything. In (d) the limit is 0 by
L’Hôpital’s rule used correctly. Only one application is enough.
5
10. Let f (u) = tan−1 (u), a function
R v whose domain in the set of all real numbers and whose
π π
range is (− 2 , 2 ). Let g(v) = 0 f (t)dt.
(a) f (1) = π4 .
(b) f (1) + f (2) + f (3) = π.
(c) g is an increasing function on the entire real line.
(d) g is an odd function, i.e., g(−x) = −g(x) for all real x.
Correct options: a,b Wrong: c,d
(a) is direct and (b) is a straightforward calculation using the formula for tan(A + B)
keeping in mind the range of tan−1 . By the fundamental theorem of calculus, g 0 (x) =
f (x), so g is increasing when f is positive, which is true only in (0, ∞). g is an
even
R q function Rasp its derivative f is odd. Note that g(x) is defined for all real x as
p
f (t)dt = − q f (t)dt.
Part B Solutions
B1. Solve the following two independent problems on pages 2–3 of the answer booklet.
(i) Let f be a function from domain S to codomain T . Let g be another function from
domain T to codomain U . For each of the blanks below choose a single letter corre-
sponding to one of the four options listed underneath. (It is not necessary that each
choice is used exactly once.) Write your answers on page 2 as a sequence of four letters
in correct order. Do NOT explain your answers.
If g ◦ f is one-to-one then f B and g D .

If g ◦ f is onto then f D and g C .
Option A: must be one-to-one and must be onto.

Option B: must be one-to-one but need not be onto.
Option C: need not be one-to-one but must be onto.
Option D: need not be one-to-one and need not be onto.
Recall: g ◦ f is the function defined by g ◦ f (a) = g(f (a)). The function f is said
to be one-to-one if, for any a1 and any a2 in S, f (a1 ) = f (a2 ) implies a1 = a2 . The
function f is said to be onto if, for any b in T , there is an a in S such that f (a) = b.
6
(ii) In the given figure ABCD is a square. Points X and Y , respectively on sides BC and
CD, are such that X lies on the circle with diameter AY . What is the area of the
square ABCD if AX = 4 and AY = 5? (Figure is schematic and not to scale.)
B X C
A D
Solution: AXY is a right angle, being an angle in a semicircle. Therefore by Pythagoras,

XY = 3. Triangles ABX and XCY are similar, because both are right angled triangles and
at the point X the three angles add to 180◦ , with the middle angle AXY being a right angle.
We have the following three equations in three unknowns.
AB 4
= by similarity BX + CX = AB AB 2 + BX 2 = 16.
CX 3
Solving these gives the answer, e.g., CX = 43 AB by the first equation, so BX = 14 AB by
1 2
the second equation, so AB 2 + 16 AB 2 = 16 by the third equation, so area = AB 2 = 16
17
.
B2. Solve the following two independent problems on pages 4–5 of the answer booklet.
(i) A mother and her two daughters participate in a game show. At first, the mother
tosses a fair coin.
Case 1: If the result is heads, then all three win individual prizes and the game ends.
Case 2: If the result is tails, then each daughter separately throws a fair die and wins
a prize if the result of her die is 5 or 6. (Note that in case 2 there are two independent
throws involved and whether each daughter gets a prize or not is unaffected by the
other daughter’s throw.)
(a) Suppose the first daughter did not win a prize. What is the probability that the
second daughter also did not win a prize?
Solution: Since the first daughter did not win a prize, the coin toss must have
shown tails. Now the second daughter does not win precisely when she throws 1,
2, 3 or 4. The probability of this is unaffected by the first daughter’s throw. So
the desired probability is 46 = 23 . One can also do this more pedantically in a way
similar to part (b), see below.
7
(b) Suppose the first daughter won a prize. What is the probability that the second
daughter also won a prize?
Solution: Let T = the event that the coin toss gives tails. Similarly H = heads,
F = first daughter wins, S = second daughter wins. We want P(S | F). Note that
the outcome of the throw of each die is independent of that of the other die and
is unaffected by the coin toss that preceded it.
P(S | F) = P(S & F)/P(F) = ( 59 )/( 23 ) = 5
6
because
1
P(F) = P(H) + P(T) P(first die = 5 or 6 | T) = 2
+ 21 26 = 23 , and
1
P(S & F) = P(H) + P(T) P(each die = 5 or 6 | T) = 2
+ 12 ( 26 )2 = 59 .
We can also solve part (a) similarly to find the desired P(not S | not F).
P(not S | not F) = P((not S) & (not F))/ P(not F) = 64 because
P(not F) = P(T) P(1 ≤ first die ≤ 4 | T) = 21 46 , and
P((not S) & (not F)) = P(T) P(1 ≤ both dice ≤ 4 | T) = 12 ( 64 )2 .
(ii) Prove or disprove each of the following statements.

(a) 240 > 20!
Solution: False. In fact 262 > 20! > 261 , so even crude estimation is enough
to solve this. For example LHS = 240 = 420 = 4 × 4 × · · · × 4 (20 times).
RHS = 1 × 2 × · · · × 20. Consider the ratio RHS/LHS and pair the 20 numbers
in each product. The initial three fractions less than 1, namely 41 , 24 , 34 are easily
overpowered by the remaining ones, e.g., they are individually matched by 16 , 8, 6.
4 4 4
(OR, using 2 × 3 > 22 , 4 × · · · × 7 > 44 = 28 , 8 × · · · × 15 > 88 = 224 and
16 × · · · × 20 > 165 = 220 , one gets 20! > 254 .)
1
(b) 1 − x
≤ ln x ≤ x − 1 for all x > 0.
Solution: True. Let f (x) = x − 1 − ln x. By analyzing the sign of f 0 (x) = 1 − x1
(or by looking at the sign of f 00 (x)), see that f (x) has a global minimum at x = 1
and that this minimum value is 0, giving ln x ≤ x − 1. For the other inequality,
substitute x = 1t in ln x ≤ x − 1 to get ln 1t = − ln t ≤ 1t − 1, i.e., 1 − 1t ≤ ln t
for all t such that 1t > 0, which is equivalent to having the same inequality for all
t > 0. So we may replace t by x, giving the desired result. (Of course, it is also
possible to repeat the earlier logic by analyzing ln x − 1 + x1 .)
B3. You are supposed to create a 7-character long password for your mobile device.
(i) How many 7-character passwords can be formed from the 10 digits and 26 letters?
(Only lowercase letters are taken throughout the problem.) Repeats are allowed, e.g.,
0001a1a is a valid password.
Solution: For each character there are 36 choices. So number of passwords = 367 .
8
(ii) How many of the passwords contain at least one of the 26 letters and at least one of
the 10 digits? Write your answer in the form: (Answer to part i) − (something).
Solution: From 367 remove 267 passwords containing only letters and 107 passwords
containing only digits. Required number = 367 − (267 + 107 ).
(iii) How many of the passwords contain at least one of the 5 vowels, at least one of the 21
consonants and at least one of the 10 digits? Extend your method for part ii to write
a formula and explain your reasoning.
Solution: Apply the inclusion exclusion principle or use a Venn diagram. Out of 367
passwords, 317 contain no vowels (V), 157 contain no consonants (C) and 267 contain no
digits (D). As first step we take 367 −(317 missing V + 157 missing C + 267 missing D).
But this removes the passwords without two types of characters (i.e., those containing
only one type of character) twice. So we need to add these back so as to effectively
remove them only once. So we need to add (107 missing VC + 217 missing VD + 57
missing CD). So the final answer is
367 − (317 + 157 + 267 ) + (107 + 217 + 57 ).
(iv) Now suppose that in addition to the lowercase letters and digits, you can also use 12
special characters. How many 7-character passwords are there that contain at least
one of the 5 vowels, at least one of the 21 consonants, at least one of the 10 digits and
at least one of the 12 special characters? Write only the final formula analogous to
your answer to part iii. Do NOT explain.
Solution: The answer is
487 − (437 + 277 + 387 + 367 ) + (227 + 337 + 177 + 317 + 157 + 267 ) + (57 + 217 + 107 + 127 ).
A password missing one type of character is subtracted only in the second term. A
password missing two types of character is subtracted twice in the second term, so
added once in the third term. A passwordmissing three types of character is subtracted
thrice in the second term, added back 32 = 3 times in the third term and subtracted
once in the last term. (This is easier to understand as an application of the inclusion
exclusion principle. Venn diagram gets harder to keep track of as there are more
possibilities for overlaps.)
B4. Show that there is no polynomial p(x) for which cos(θ) = p(sin θ) for all angles θ in
some nonempty interval.
Hint: Note that x and |x| are different functions but their values are equal on an interval
(as x = |x| for all x ≥ 0). You may want to show as a first step that this cannot happen for
two polynomials, i.e., if polynomials f and g satisfy f (x) = g(x) for all x in some interval,
9
then f and g must be equal as polynomials, i.e., in each degree they must have the same
coefficient.
Solution: To prove the assertion in the hint, note that the polynomial f − g would have
infinitely many roots and hence must be the zero polynomial.
Suppose a polynomial p satisfies cos(θ) = p(sin θ) for θ in an interval. Let t = sin θ. Then
the following equality is true for the (infinitely many) values of t in some nonempty interval:
p(t)2 = cos2 (θ) = 1 − sin2 (θ) = 1 − t2 .
By the hint, this forces the polynomial 1 − x2 to be the square of the polynomial p(x). But
1 − x2 is not a square because, e.g., the square of the leading coefficient would need to be
−1, which cannot happen. (Or p would need to be a linear polynomial ax + b, etc.)
Note: The italicized part “in some nonempty interval” at the end of the problem statement
was missing in the actual exam due to oversight. As sin of two angles can be equal without
their cos being equal, a very easy solution now becomes possible, e.g., just plug in θ = 0 and
θ = π to get p(0) = 1 as well as p(0) = −1. So in fact there cannot be any function (not
just polynomial) p such that cos(θ) = p(sin θ) for ALL angles θ. In the exam, everyone who
gave any correct solution to the problem as stated there was given full credit.
B5. Define a function f as follows: f (0) = 0 and, for any x > 0,

Z L Z ∞
1 1
f (x) = lim cos(t)dt or, in simpler notation, the improper integral cos(t)dt .
L→∞ 1 t2 1 t2
x x
(i) Show that the definition makes sense for any x > 0 by justifying why the limit in the
definition exists, i.e., why the improper integral converges.
Solution: As | cos(t)| ≤ 1, the integral defining f (x) is in fact absolutely convergent.
Z ∞ Z ∞
1 1 1 ∞
2
cos(t) dt ≤ 2
dt = − = x.
1
x
t 1
x
t t 1
x
(ii) Find f 0 ( π1 ) if it exists. Clearly indicate the basic result(s) you are using.
df
Solution: For x 6= 0, let u = x1 . By the fundamental theorem of calculus1 , du = − u12 cos(u) =
df
−x2 cos( x1 ). Since du
dx
= − x12 , by chain rule dx = cos( x1 ), so f 0 ( π1 ) = cos(π) = −1.
(iii) Using the hint or otherwise, find limh→0+ f (h)−f
h
(0)
, i.e., the right hand derivative of f
at x = 0. We can take the limit only from the right hand side because f (x) is undefined for
negative values of x.
Hint: Break f (h) into two terms by using a standard technique with an appropriate choice.
Then separately analyze the resulting two terms in the derivative.
1
To use the standard version of the fundamental theorem where the lower endpoint is a fixed finite
R1
number, take some positive constant K. Then f (x) = − Kx t12 cos(t)dt + a constant (by part i).
10
Solution: Integrate by parts using u = t12 and dv = cos(t)dt. Then du = − t23 dt and
v = sin(t). (The method below does not work with u = cos(t) and dv = t12 dt.) So
f (h) − f (0) 1 ∞ 1 1 sin(t) ∞ 1 ∞ 2 sin(t)

Z Z
= cos(t)dt = + dt.
h h h1 t2 h t2 h1 h h1 t3
∞
1 sin(t)
The first term in the sum = h t2 1
= −h sin( h1 ) → 0 as h → 0+ , where we have twice used
h
that | sin(t)| ≤ 1 for all t. For the the second term, by logic similar to part (i) we get
1 ∞ 2 sin(t) 1 ∞ 2
Z Z
1 −2 ∞
dt ≤ dt = − t 1 = h → 0 as h → 0+ .
h h1 t3 h h1 t3 h h
So the desired limit is 0.

Rx
Note: Substituting s = 1t gives f (x) = the simpler looking improper integral 0 cos( 1s )ds.
This makesR x parts (i) and (ii) more transparent. Now x < 0 is also ok in the improper
1
integral 0 cos( s )ds. For the function g(x) defined by this new integral, the above analysis
gives one way to show that g 0 (0) exists (see math.stackexchange.com/questions/2127903). g
is differentiable everywhere (the only case requiring work being x = 0) but the derivative is
not continuous at x = 0, as g 0 does not have a limit at x = 0.
B6. n and k are positive integers, not necessarily distinct. You are given two stacks of cards
with a number written on each card, as follows.
Stack A has n cards. On each card a number in the set {1, . . . , k} is written.
Stack B has k cards. On each card a number in the set {1, . . . , n} is written.
Numbers may repeat in either stack. From this, you play a game by constructing a sequence
t0 , t1 , t2 , . . . of integers as follows. Set t0 = 0. For j > 0, there are two cases:
If tj ≤ 0, draw the top card of stack A. Set tj+1 = tj + the number written on this card.
If tj > 0, draw the top card of stack B. Set tj+1 = tj − the number written on this card.
In either case discard the taken card and continue. The game ends when you try to draw
from an empty stack. Example: Let n = 5, k = 3, stack A = 1, 3, 2, 3, 2 and stack B = 2, 5, 1.
You can check that the game ends with the sequence 0, 1, −1, 2, −3, −1, 2, 1 (and with one
card from stack A left unused).
(i) Prove that for every j we have −n + 1 ≤ tj ≤ k.
(ii) Prove that there are at least two distinct indices i and j such that ti = tj .
(iii) Using the previous parts or otherwise, prove that there is a nonempty subset of cards
in stack A and another subset of cards in stack B such that the sum of numbers in
both the subsets is same.
11
Solution: (i) Induction. Base case is true as t0 = 0. Assume the result up to tj . Now there
are two cases. If tj ∈ [−n + 1, 0] then tj+1 = tj + a number from stack A, which is between
1 and k, so tj+1 ∈ [(−n + 1) + lowest possibility 1, 0 + highest possibility k] ⊂ [−n + 1, k].
If tj ∈ [1, k] then tj+1 = tj − a number from stack B, which is between 1 and n, so tj+1 ∈
[1 − highest possibility n, k − lowest possibility 1] ⊂ [−n + 1, k].
(ii) Suppose the game ends when we try to draw from stack B. As there are k cards in stack
B, in all we must have tried to draw k + 1 times from stack B. At each one of these attempts,
the value of tj must have been positive and by part (i) each one of these k + 1 numbers is
between 1 and k (inclusive). So there must be a repeat among these k + 1 numbers.
If the game ends when we try to draw from stack A, the argument is parallel: there must
have been n + 1 attempts to draw from stack A, each one resulting from a value of tj that
lies among the n numbers from −(n − 1) to 0, so there must be a repeat.
(iii) Suppose ti = tj . Then, from the set of cards drawn at steps i + 1, . . . , j the sum of the
cards from stack A must equal the sum of the cards from stack B.
12
CHENNAI MATHEMATICAL INSTITUTE
Undergraduate Programme in Mathematics and Computer Science/Physics
Solutions of the 22nd May 2022 exam
Note: The solutions below consist only of main steps and strategies and do not contain all
the details expected in the exam.
B1. [11 points] Given △XY Z, the following constructions are made: mark point W on
segment XZ, point P on segment XW and point Q on segment Y Z such that
WZ PW QZ
= = = k.
YX XP YQ
See the schematic figure (not to scale). Extend segments QP and Y X to meet at the point
R as shown. Prove that XR = XP .
R
P W
Y Z
Q
Solution: First a construction - mark V on XZ such that QV is parallel to Y R. There are

two cases here depending on whether V is between P W or W Z, however, the arguments are
the same. We assume here that V is between P W . The aim is to show that △V P Q is isosceles
k
and then show that it is similar to △XP R. Use BPT to conclude that V Q = k+1 (XY ).
Using the given ratios find an expression for V Z and substitute it in P V = P Z − V Z to
conclude that P V = V Q.
One can also extend ZX to ZX ′ such that Y X ′ is parallel to P Q. One can then show that
△Y XX ′ is isosceles and similar to △RXP .
Another strategy is to use Menalaus theorem for △XY Z with segment QR as the transversal.
We have:
XR Y Q ZP
= −1.
RY QZ P X
This leads to the following implications leading to the equality we want:
XR · P Z = RY · P W
XY + XR PW + WZ
=
XR PW
XY WZ
=
XR PZ
PW PW
= .
XR XP
1
B2. [11 points] In the XY plane, draw horizontal and vertical lines through each integer on
both axes so as to get a grid of small 1 × 1 squares whose vertices have integer coordinates.
1. Consider the line segment D joining (0, 0) with (m, n). Find the number of small 1 × 1
squares that D cuts through, i.e., squares whose interiors D intersect. For example,
the line segment joining (0, 0) and (2, 3) cuts through 4 small squares.
2. Now let L be an arbitrary line. Find the maximum number of small 1 × 1 squares in
an n × n grid that L can cut through.
Solution: Assume gcd(m, n) = 1. The line D has to cross m−1 vertical as well as horizontal
lines. Moreover, D doesn’t pass through any grid points. Hence, together with the starting
square, we see that D cuts through m + n − 1 squares.
Let gcd(m, n) = d. The above argument is valid from (0, 0) to (m/d, n/d) and so on for d
many sections of D. Therefore the total number of squares D cuts is m + n − d.
Note that in order for L to cut through maximum number of squares it should not pass
through any internal grid point. This is possible for a line joining (0, 0) with (x, n) where
n − 1 < x < n. The required answer is 2n − 1.
2
B3. [14 points] For a positive integer n, let f (x) := 1 + x + x2 · · · + xn . Find the number
of local maxima of f (x). Find the number of local minima of f (x). For each maximum/
minimum (c, f (c)), find the integer k such that k ≤ c < k + 1.
Solution: We have f ′ (x) = 1 + 2x + · · · + nxn−1 . For x ≥ 0 the derivative is strictly positive,
hence f (x) is strictly increasing. Therefore, we should only analyze negative values of x.
Write the derivative as the following rational function
nxn+1 − (n + 1)xn + 1
f ′ (x) = .
(x − 1)2
Note that there is no problem in the expression since we are assuming x < 0. Denote by
D(x) the denominator of the derivative.
The case when n is odd. For x < 0 the polynomial D(x) is strictly positive. Hence there
can’t be any critical point.
The case when n is even. Observe that there could be only one critical point c ∈ (−1, 0).
Since D(x) < 0 for x ≤ −1 and D(0) = 1. Moreover, D′ (x) > 0 for x < 0 so f ′ (x)
is increasing on (−∞, 0) hence it vanishes exactly once. As the derivative changes sign
from −ve to +ve passing through c, so there is exactly one global minimum at c (where,
−1 < c < 0).
3
B4. [14 points] For a continuous function f : R+ → R+ , define
• Ar = the area bounded by the graph of f , X-axis, x = 1 and x = r.
• Br = the area bounded by the graph of f , X-axis, x = r and x = r2 .
Find all continuous functions f for which Ar = Br for every positive number r.
Solution: We are given
Z r Z r2
f (x)dx = f (x)dx.
1 r
Applying d/dr, fundamental theorem of calculus and the chain rule to above equality we get
f (x) = xf (x2 ) ∀x ∈ R+ .
1
Letting g(x) = xf (x) we see that g(x) = g(x2 ) for all x in the domain. Hence g(x) = g(x 2n )
1
for all x and positive integers n. However, as n goes to infinity x 2n tends to 1 we have that
g(x) converges to f (1). Hence xf (x) = f (1) for all values of x ∈ R+ .
4
B5. [14 points] Two distinct real numbers r and s are said to form a good pair (r, s) if
r 3 + s2 = s3 + r 2 .
1. Find a good pair (a, l) with the largest possible value of l. Find a good pair (s, b) with
the smallest value of s. For every good pair (c, d) other than the two you found, show
that there is a third real number e such that (d, e) and (c, e) are good pairs.
2. Show that there are infinitely many good pairs of rational numbers.
Solution: Consider the function f (x) = x3 − x2 . Therefore (r, s) is a good pair iff f (r) =
f (s).
Observe that x = 0, 32 are the only critical points of f . The local maximum occurs at x = 0.
The line y = 0 intersects the graph of f (x) at (0, 0) and (1, 0). Hence the required good pair
(a, l) with the largest l value is (0, 1).
Note that the local minimum occurs at x = 32 . The line y = f ( 23 ) = −427
intersects the graph
−1 −4 2 −4 −1 2
at ( 3 , 27 ) and ( 3 , 27 ). Hence the required good pair is ( 3 , 3 ).
For k ∈ ( −427
, 0) the line y = k intersects the graph at 3 points. Hence the last statement of
the first part follows.
For the second part we need to show that there for every rational number q ∈ ( −4 27
, 0) the
equation
x3 − x2 − q = 0
has infinitely many rational solutions. However, this is true because there are infinitely many
rationals satisfying c + d + e = 1, cd + de + ce = 0, cde = q.
5
B6. [14 points] Solve the following.
1. Let p be a prime. Show that x2 + x − 1 has at most two roots modulo p. Find all
primes p for which there is exactly one root.
2. Find all positive integers n ≤ 121 such that n2 + n − 1 is divisible by 121.
3. What can you say about the number of roots of this equation modulo p2 .
Solution: Let a, b be two distinct roots of the equation modulo p. Therefore, p divides
a2 + a − 1 − (b2 + b − 1), which is equivalent to saying that p divides either a − b or a + b + 1.
In the former case we will have a = b, which is not allowed. Since both a, b are between 1
and p we have 3 ≤ a + b + 1 ≤ 2p − 3 which implies a + b + 1 = p. Thus b = p − a − 1 is
uniquely determined.
Suppose a is the only root. Then p − a − 1 = a, i.e., p = 2a + 1. Therefore, 2a + 1 divides
4(a+ a − 1) and (2a + 1)2 . Subtracting we get that 2a + 1 divides 5.
Part 2: Since 121 divides n+ n − 1, 11 also divides it. Note that n2 + n − 1 and n2 + n − 12
are congruent modulo 11. So the roots of the equation are 7, 3 modulo 11.
Consider n = 3 + 11k. Then n2 + n − 1 is congruent to 77k + 11 modulo 121. Then k = 3
works giving us n = 36. Now consider n = 7 + 11k. In that case, n+ n − 1 is congruent to
165k + 55 modulo 121. Which gives us k = 7 and n = 84.
For part (3), let a be a root modulo p. Then n is of the form kp + a for some k between 0
and p − 1. We would like to solve for k the following equation
(kp + a)2 + (kp + a) − 1
modulo p2 . This is equivalent to finding k such that p divides k(2a + 1) + a2 + a − 1. If

2a + 1 is not a multiple of p then k = −(2a + 1)−1 (a2 + a − 1). If p divides (2a + 1) then it
is 5 and there is no such n.
6
2023 CMI BSc entrance exam solutions (draft)
Part A
A1. Define the right derivative of a function f at x = a to be the following limit if it exists.
f (a + h) − f (a)
lim+ , where h → 0+ means
h→0 h
h approaches 0 only through positive values.
Statements
(1) If f is differentiable at x = a then f has a right derivative at x = a.
(2) f (x) = |x| has a right derivative at x = 0.
(3) If f has a right derivative at x = a then f is continuous at x = a.
(4) If f is continuous at x = a then f has a right derivative at x = a.
(1) True. Obvious from the definition of the derivative.

(2) True. Right derivative is 1.
(3) False. Consider the floor function at integer values.
(4) False. Take x sin x1 made continuous at 0.
A2. Suppose a rectangle EBF D is given and a rhombus ABCD is inscribed in it so that
the point A is on side ED of the rectangle. The diagonals of ABCD intersect at point G.
See the indicative figure below.
Statements
(5) Triangles CGD and DF B must be similar.
AC EB
(6) It must be true that BD = ED .
(7) Triangle CGD cannot be similar to triangle AEB.
(8) For any given rectangle EBF D, a rhombus ABCD
as described above can be constructed.
(5) True. Both are right angled and ∠F BD = ∠BDA = ∠GDC.

(6) True. G bisects AC and BD. Use the similarity of CGD with DEB ' DF B.
(7) False. They are similar when ABC is equilateral, which is possible.
(8) False. ED ≥ AD = AB ≥ EB (hypotenuse), so need ED ≥ EB.
A3. This question is about complex numbers.
Statements
(9) The complex number (e3 )i lies in the third quadrant.
(10) If |z1 | − |z2 | = |z1 + z2 | for some complex numbers z1 and z2 , then z2 must be 0.
(11) For distinct complex numbers z1 and z2 , the equation |(z − z1 )2 | = |(z − z2 )2 | has at
most 4 solutions.
(12) For each nonzero complex number z, there are more than 100 numbers w such that
w2023 = z.
(9) False. Second quadrant. The argument of e3i is 3 radian, which is just under 172◦ .
(10) False. Take z2 = rz1 with r real and −1 ≤ r < 0.
(11) False. |(z − z1 )2 | = |z − z1 |2 , so z is equidistant from z1 and z2 . Solutions form a line.
1 i(2πk+θ)
z
(12) True. There are 2023 such w. Letting |z|
= eiθ , w = |z| 2023 e 2023 , k = 0, 1, . . . , 2022.
A4. Statements
(13)
1
lim e x = +∞.
x→0
(14)
ln x ln x
lim < lim 1 .
x→∞ x100 x→∞ x 100
(15) For any positive integer n,

Z n
n
x2023 cos(nx)dx < .
−n 2023
(16) There is no polynomial p(x) for which there is a single line that is tangent to the graph
of p(x) at exactly 100 points.
(13) False. Limit from the left is 0.

(14) False. Both limits are 0. The numerator ln x is dominated by any positive power of x.
(15) True. The function x2023 cos(nx) is odd so by symmetry the integral is 0.
(16) False. For p(x) with exactly 100 multiple roots the X-axis is such a line. (This is
essentially the only way: if y = ax + b is such a line for a polynomial q(x), then the X-axis
is such a line for the polynomial q(x) − ax − b, which must have exactly 100 multiple roots.)
A5. Statements
p √
(17) 4 < 5 + 5 5.
1+log2 61
(18) log2 11 < 2
.
(19) (2023)2023 < (2023!)2 .
(20) 92100 + 93100 < 94100 .
√ √
(17) True. 16 < 5 + 5 5 as 11 < 5 5 as 121 < 125. Taking square roots preserves order.
(18) True. 22 log2 11 = 121 < 21+log2 61 = 122. Taking log2 preserves order.
(19) True. Pair numbers on the RHS symmetrically. n(2024 − n) > 2023 for n = 1, . . . , 2023.
1 2
(20) True. Divide by 92100 and use binomial theorem for (1 + x)100 with x = 92
and x = 92
.
P n
P
A6. For a sequence ai of real numbers, we say that ai converges if lim ai is finite.
n→∞ i=1
In this question all ai > 0.
Statements
P
(21) If ai converges, then ai → 0 as i → ∞.
(22) If ai < 1i for all i, then
P
ai converges.
ai converges, then (−1)i ai also converges.
P P
(23) If
P P
(24) If ai does not converge, then i tan(ai ) cannot converge.
(21) True.
1
(22) False. Take ai = 2i
n
(−1)i ai remains a Cauchy sequence.
P
(23) True. Given ai are all positive, so
i=1
(24) False. Take all ai = π.
Continued −→
A7. Statements
(25) To divide an integer b by a nonzero integer d, define a quotient q and a remainder r to
be integers such that b = qd + r and |r| < |d|. Such integers q and r always exist and are
both unique for given b and d.
(26) To divide a polynomial b(x) by a nonzero polynomial d(x), define a quotient q(x) and
a remainder r(x) to be polynomials such that b = qd + r and degree(r) < degree(d). (Here
b(x) and d(x) have real coefficients and the 0 polynomial is taken to have negative degree
by convention.) Such polynomials q(x) and r(x) always exist and are both unique for given
b(x) and d(x).
(27) Suppose that in the preceding question b(x) and d(x) have rational coefficients. Then
q(x) and r(x), if they exist, must also have rational coefficients.
(28) The least positive number in the set {(a × 20232020 ) + (b × 20202023 )} as a and b range
over all integers is 3.
(25) False. For r > 0, we can increase quotient by 1 and make remainder negative.
(26) True. q1 d + r1 = q2 d + r2 gives (q1 − q2 )d = r2 − r1 . Compare degrees.
(27) True. Uniqueness and the long division procedure ensure this.
(28) False. It is the gcd of 20232020 and 20202023 , which is 1.
A8. You play the following game with three fair dice. (When each one is rolled, any one of
the outcomes 1, 2, 3, 4, 5, 6 is equally likely.) In the first round, you roll all three dice. You
remove every die that shows 6. If any dice remain, you roll all the remaining dice again in
the second round. Again you remove all dice showing 6 and continue.
Questions
(29) Let the probability that you are able to play the second round be ab , where a and b are
integers with gcd 1. Write the numbers a and b separated by a comma. E.g., for probability
10
36
you would type 5,18 with no quotations, space, full stop or any other punctuation.
(30) Let the probability that you are able to play the second round but not the third round
be dc where c and d are integers with gcd 1. Write only the integer c as your answer. E.g.,
34
for probability 36 you would type 17 with no quotations, space or any other punctuation.
(29) The probability is 1 − ( 61 )3 = 215
216
, so 215,216 is the answer.
1115
(30) The probability is 66
by the calculation below, so 1115 is the answer.
P(3 dice left after first round) × P(all 3 remaining dice show 6 in round 2) +
P(2 dice left after first round) × P(both remaining dice show 6 in round 2) +
P(1 die left after first round) × P(the remaining die shows 6 in round 2)
5 3 1 3 5 2 1 1 2 5 1 2 1
= × + 3 × + 3 × .
6 6 6 6 6 6 6 6
A9. Two lines `1 and `2 in 3-dimensional space are given by
`1 = {(t − 9, −t + 7, 6) | t ∈ R} and `2 = {(7, s + 3, 3s + 4) | s ∈ R}.
Questions
(31) The plane passing through the origin and not intersecting either of `1 and `2 has equation
ax + by + cz = d. Write the value of |a + b + c + d| where a, b, c, d are integers with gcd = 1.
(32) Let r be the smallest possible RADIUS of a circle that has a point on `1 as well as a
point on `2 . It is given that r2 (i.e, the SQUARE of the smallest radius) is an integer. Write
the value of r2 .
(31) (1, −1, 0) and (0, 1, 3) are the direction vectors of the given lines. (3, 3, −1) is a common
perpendicular to both direction vectors. So 3x + 3y − z = 0 is an equation for the desired
plane. Thus the answer is |3 + 3 − 1 + 0| = 5.
(32) Each of the two mentioned points must be the only intersection of such a circle with
the respective line. The segment joining these points must be perpendicular to both `1 and
`2 and is a diameter of any specified circle. Taking a general point on each line, a vector
representing the segment joining the two points is (16 − t, s + t − 4, 3s − 2). Solving
(16 − t, s + t − 4, 3s − 2) · (1, −1, 0) = 0 and (16 − t, s + t − 4, 3s − 2) · (0, 1, 3) = 0
gives s = 0, t = 10. So (2r)2 = (16 − 10)2 + (10 − 4)2 + (−2)2 = 76. Thus the answer is 19.
A10. Consider the part of the graph of y 2 + x3 = 15xy that is strictly to the right of the
Y-axis, i.e., take only the points on the graph with x > 0.
Questions
(33) Write the least possible value of y among considered points. If there is no such real
number, write NONE (without any spaces or quotation marks or any other punctuation).
(34) Write the largest possible value of y among considered points. If there is no such real
number, write NONE (without any spaces or quotation marks or any other punctuation).
√
(33) Regarding the equation as a quadratic in y gives y = 12 15x ± 225x2 − 4x3 . So

there is a y-value for every x ≤ 225 . We also have y > 0 for x > 0 since in that case
√ 4
2 3
15x > 225x − 4x . As x → 0, y also → 0, so there is no minimum y-value.
(34) On the closed interval [0, 225

4
], each sign in the formula for y gives a continuous function.
So y must have a maximum, which cannot occur at x = 0 as y > 0 for x > 0. So we inspect
y-values for the endpoint x = 225 4
and for any critical points. Differentiating implicitly,
dy 2
dx
= 0 precisely when 3x = 15y for (x, y) on the graph. This gives only the point (50, 500).
As 500 > (y-value at x = 225 4
), the answer is 500. One can also argue without considering
225
x = 4 at all. Plot the graph to see that there is no “endpoint”.
Part B solutions (draft)
B1. [11 points] We want to find odd integers n > 1 for which n is a factor of 2023n − 1.
(a) Find the two smallest such integers.
(b) Prove that there are infinitely many such integers.
Solution: (a) 2023 is 1 mod 3, so n = 3 works. Similarly using modular arithmetic one
checks that 5 and 7 do not work but 9 does. (b) If n = k works so does n = 3k by induction:
20233k −1 = (2023k −1)(20232k +2023k +1) = (multiple of k)(multiple of 3) as each summand
in the second factor is 1 mod 3. Thus all powers of 3 satisfy the required condition.
B2. [12 points] Let Z+ denote the set of positive integers. We want to find all functions
g : Z+ → Z+ such that the following equation holds for any m, n in Z+ .
g(n + m) = g(n) + nm(n + m) + g(m).
Prove that g(n) must be of the form di=0 ci ni and find the precise necessary and sufficient
P
condition(s) on d and on the coefficients c0 , . . . , cd for g to satisfy the required equation.
Solution: Setting m = 1 gives g(n + 1) − Pg(n) = n(n + 1) + g(1). Apply this repeatedly
n−1 2
starting with n = 1 to get g(n) = ng(1) + i=1 i + i, which works out to ng(1) + 31 (n3 − n).
Set g(1) to be an arbitrary positive integer k and verify that the resulting formula satisfies
the given condition for all m, n. Thus g(n) = 31 n3 + kn with k = (any positive integer)− 31 .
B3. [13 points] Suppose that for a given polynomial p(x) = x4 + ax3 + bx2 + cx + d, there
is exactly one real number r such that p(r) = 0.
(a) If a, b, c, d are rational, show that r must be rational.
(b) If a, b, c, d are integers, show that r must be an integer.
Possible hint: Also consider the roots of the derivative p0 (x).
Solution: (a) The multiplicity of the root r must be either 2 or 4. In the latter case
p(x) = (x − r)4 = x4 − 4rx3 + 6r2 x2 − 4r3 x + r4 . As 4r is a rational, so is r. If the multiplicity
is 2 then the two non-real roots are complex conjugates and r is the only repeated root
of p(x). So r is the only common root (whether real or not) of p(x) and p0 (x) AND r
is a simple root of the polynomial p0 (x). So gcd(p, p0 ) = x − r by looking at complete
factorization of p(x) and p0 (x) into linear terms (including complex roots). As p(x) and p0 (x)
have rational coefficients, so does their gcd by looking at each step of the division algorithm.
(The preceding three sentences deserve careful consideration.) Therefore r must be rational.
(b) It is standard that a rational root of a monic polynomial with integer coefficients must
be an integer. (Proof: write a rational root r = mn
with gcd(m, n) = 1, substitute into the
polynomial, clear powers of n in the denominators and deduce that n cannot be divisible by
any prime because that prime would then need to divide m as well. So n = ±1.)
B4. [14 points] There are n students in a class and no two of them have the same height.
The students stand in a line, one behind another, in no particular order of their heights.
(a) How many different orders are there in which the shortest student is not in the first
position and the tallest student is not in the last position?
(b) The badness of an ordering is the largest number k with the following property. There
is at least one student X such that there are k students taller than X standing ahead
of X. Find a formula for gk (n) = number of orderings of n students with badness k.
Example: The ordering 64 61 67 63 62 66 65 (the numbers denote heights) has badness

3 as the student with height 62 has three taller students (with heights 64, 67 and 63)
standing ahead in the line and nobody has more than 3 taller students standing ahead.
Possible hints for (b): It may be useful to first count orderings of badness 1 and/or to find
fk (n) = the number of orderings of n students with badness less than or equal to k.
Solution: (a) There are (n − 1)! + (n − 1)! − (n − 2)! orderings with the shortest student
first or the tallest student last or both. So the desired number = n! − 2(n − 1)! + (n − 2)!,
i.e., (n − 2)!(n2 − 3n + 3). Alternatively, first order all but the shortest and the tallest
students in (n − 2)! ways. The number of ways to insert the shortest and then the tallest is
(n − 2)(n − 1) + 1. (What is the extra 1 for?)
(b) Following both the hints, first consider badness 1 and use induction. Leave out the
shortest student and order the remaining n−1 students with badness 1. The shortest student
can now go in place 1 or 2. There is one more possibility where the n − 1 students have 0
badness (i.e., are in increasing order) and the shortest student goes in place 2. Inductively
one gets the formula g1 (n) = 2n−1 − 1 (valid even for n = 1, giving g1 (1) = 0).
Induction to find fk (n) is easier. Leave out the shortest student and order the remaining
n − 1 students with badness at most k. To maintain badness at most k, out of the n
available slots for the shortest student, the allowed ones are precisely 1, 2, . . . , min(k + 1, n).
So fk (n + 1) = fk (n) min(k + 1, n). Answer: fk (n) = n! if n ≤ k + 1 and fk (n) = k!(k + 1)n−k
if n ≥ k + 1. (The formulas agree for n = k + 1.)
Now gk (n) = fk (n) − fk−1 (n). This works out to 0 if n ≤ k (as expected) and for n ≥ k, one
gets gk (n) = k! (k + 1)n−k − k n−k .
B5. [15 points] Throughout this question every mentioned function is required to be a
differentiable function from R to R. The symbol ◦ denotes composition of functions.
(a) Suppose f ◦ f = f . Then for each x, one must have f 0 (x) = or f 0 (f (x)) = .
Complete the sentence and justify.
(b) For a non-constant f satisfying f ◦ f = f , it is known and you may assume that the
range of f must have one of the following forms: R, (−∞, b], [a, ∞) or [a, b]. Show that
in fact the range must be all of R and deduce that there is a unique such function f .
(Possible hints: For each y in the range of f , what can you say about f (y)? If the
range has a maximum element b what can you say about the derivative of f ?)
(c) Suppose that g ◦ g ◦ g = g and that g ◦ g is a non-constant function. Show that g must
be onto, g must be strictly increasing or strictly decreasing and that there is a unique
such increasing g.
Solution: (a) f 0 (f (x))f 0 (x) = f 0 (x) for each x by chain rule, so f 0 (x) = 0 or f 0 (f (x)) = 1.
(b) (Argument taken from the answer by Dan Schved to question 365363 on stackexchange.)
For each y = f (x) in the range, f (y) = f (f (x)) = f (x) = y, so f is the identity function
on the range. Therefore it is enough to show that the range is all of R. As the range is
given to be an interval (a proof is given below), at each y in the range, f 0 (y) = 1 by direct
calculation. Note that if the range has endpoint(s), this derivative calculation is one sided
at such a point. We will show that the range does not have an endpoint on either side. If the
range interval has a left/right endpoint f (c) = c, then f has a minimum/maximum value at
c, so it must be true that f 0 (c) = 0. (Recall that the domain is all of R, so Fermat’s theorem
applies at x = c.) This contradicts the earlier calculation of a one-sided derivative at c being
1. So the range cannot be of the form (−∞, b], [a, ∞) or [a, b] and must be all of R.
(c) By applying g to the given equation, g ◦ g ◦ g ◦ g = g ◦ g. So if g ◦ g is non-constant, it
has to be the identity by part (b), i.e., g is its own inverse. In particular, being invertible,
g is onto and one-to-one. Due to continuity, being one-to-one implies that g is monotonic.
(This is standard. If g is one-to-one and not monotonic, we have some a, b, c for which
a < b < c and WLOG g(a) < g(c) < g(b) by replacing g with its reflection in one/both axes
if necessary. By the intermediate value theorem, we have d ∈ (a, b) with g(c) = g(d), giving a
contradiction to g being one-to-one.) If g is increasing, g(x) < x implies x = g(g(x)) < g(x)
and vice versa, so the only possibility is g(x) = x.
Proof of the fact that you were asked to assume in part (b), namely that the range of f must
be of the form R, (−∞, b], [a, ∞) or [a, b]: As f is continuous, if f (p) < r < f (q) then r is
also in the range by intermediate value theorem. So the range must be one of the intervals
(m, M ) or (m, M ] or [m, M ) or [m, M ], where m is the greatest lower bound of the range
(possibly m = −∞), M is the least upper bound of the range (possibly M = ∞), and it is
understood that if m and/or M is not finite then only the open interval makes sense on the
corresponding side. It remains to show that if either of m and M is finite, it must belong
to the range. Now f (m + h) = m + h for small enough h > 0. (Recall that the function is
non-constant so m 6= M .) So lim+ f (m + h) = lim+ m + h = m. At the same time, because
h→0 h→0
f is continuous, the same limit must be f (m), so f (m) = m and thus m is in the range of
f . For M , take limit from the left lim− f (M + h), etc.
h→0
B6. [15 points] Starting with any given positive integer a > 1 the following game is played.
If a is a perfect square, take its square root.
√ Otherwise take a + 3. Repeat the procedure
with the new positive integer (i.e., with a or a + 3 depending on the case). The resulting
set of numbers is called the trajectory of a. For example the set {3, 6, 9} is a trajectory: it
is the trajectory of each of its members.
Which numbers have a finite trajectory? Possible hint: Find the set
{n | n is the smallest number in some trajectory S}.
If you wish, you can get partial credit by solving the following simpler questions.
(a) Show that there is no trajectory of cardinality 1 or 2.
(b) Show that {3, 6, 9} is the only trajectory of cardinality 3.
(c) Show that for any integer k ≥ 3, there is a trajectory of cardinality k.
(d) Find an infinite trajectory.
√
Solution: Let S = a trajectory, n = the smallest number in S. Note that 1 ∈
/ S, so n > n.
√ √
(a) |S|= 1 implies n = n, so n = 1, which is impossible. As n > n, n cannot be a perfect
square. So |S| = 2 implies S = {n, n + 3} and n + 3 = n2 , which cannot happen for n > 1.
(b) Similarly |S| = 3 implies S = {n, n + 3, n + 6} and n + 6 = n2 , which gives n = 3.
(c) To get any finite cardinality repeatedly square 6 (or 9) and add these numbers to {3, 6, 9}.
(d) (3k)2 = 9k 2 is 0 mod 3. Next, (3k + 1)2 = 9k 2 + 6k + 1 and (3k + 2)2 = 9k 2 + 12k + 4 are
1 mod 3. As all squares are 0 or 1 mod 3, any S containing a 2 mod 3 number is infinite.
Claim: For a trajectory S with smallest number n, exactly one of the following two happens.
1. No square occurs after n in the trajectory. Hence n is 2 mod 3 and S is infinite.
2. A square does occur after n and n = 3. Hence S is finite.
Proof of the claim: The smallest number in S cannot be a square, so let k 2 < n < (k + 1)2 .
Assuming a square occurs after n, we will show that n = 3. The first encountered square
after n is at most (k + 3)2 (e.g., make cases depending on what k and n are mod 3.) So
k 2 < n ≤ k + 3, but k 2 < k + 3 only for k = 1, 2. Hence n < (k + 1)2 ≤ 9. Now n cannot be
2, 5, 8 because adding 3 repeatedly to these will never give a square. And n cannot be 4, 6, 7
because in each case one gets a smaller number by playing the game (respectively 2, 3, 2). So
n = 3 is the only possibility. For the second sentence in case 1, note that repeatedly adding
3 to n will eventually give a square if and only if n is 0 or 1 mod 3.
Main answer: If a number in S is / is not divisible by 3, then the same is true for all numbers
in S (check this). If the initial number a is a multiple of 3, then so is n, and hence we must
be in case 2 of the claim. If a is not a multiple of 3, then nor is n, so n 6= 3 and we must be
in case 1. Thus multiples of 3 are precisely the numbers with finite trajectories.
Notes: (1) The above pattern was discovered earlier by Stephan Wagner. See problem 1 in
IMO 2017 for a slightly different formulation. (2) The analysis in the solution generalizes
naturally if 3 is replaced in the game by any prime p. (Why prime?) What happens for
p = 2? For p = 5? For p = 7? In general?

CMI

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

CMI

Uploaded by

Copyright:

Available Formats

Solutions to 2015 Entrance Examination for BSc Programmes at CMI

What is the total amount donated by the 10 people?

4. Let A, B and C be unknown constants. Consider the function f (x) defined by

f (x) = Ax2 + Bx + C when x ≤ 0 ,

5. Consider the polynomial p(x) = (x + a1 )(x + a2 ) · · · (x + a10 ) where ai is a real number

q(x) = e−1/x when x > 0 ,

2. Let p, q and r be real numbers with p2 + q 2 + r2 = 1.

(a) Prove the inequality 3 p2 q + 3 p2 r + 2 q 3 + 2 r3 ≤ 2.

(b) Also find the smallest possible value of 3 p2 q + 3 p2 r + 2 q 3 + 2 r3 . Specify exactly

f 0 (x) − f 0 (y) ≤ 3|x − y|.

i The mean value theorem tells us S ⊂ T .

P (both answers correct) =

P (well prepared and both correct) 2/7 20

Solution. A bubble at the point P = (a, a2 ) must be tangential to the parabola at

3. Consider the function f (x) = xcos(x)+sin(x) defined for x ≥ 0.

(a) Prove that Z 1

(Remark: We are using calculus to analyze what in reality is a discrete situation, as a

(a) Show that |A + A| ≥ 2n − 1. Here |A + A| means the number of elements in A + A.

Solution. Suppose a polynomial f (x) leaves a constant remainder r when divided by

and solve for constants A and B to ensure desired values

As the prime p is greater than n2 , it cannot divide any of n − 1, n, n + 1. So p must divide

Read the instructions on the front of the booklet carefully!

Part A. Write your final answers on page 3.

Solution. 400 = 52 × 24 has (2 + 1) × (4 + 1) = 15 factors, so total number of pairs

5. Find all complex solutions to the equation:

Solution. It is easy to see that x4 + x3 + 2x2 + x + 1 = (x2 + 1)(x2 + x + 1).

(a) If g 00 (x0 ) = 0 then g has an inflection point at x0 . True or False?

7. Write the values of the following.

f (n) = the smallest prime factor of n.

For example, f (12) = 2, f (105) = 3. Calculate the following.

(a) limx→∞ f (x).

(a) Find the value of a for which f is continuous.

Solution. cos( x1 ) is sandwiched between −1 and 1, so limx→0 f (x) = 0 = a makes

1. Answer the following questions

First consider the limit

Now consider the following limit

1 + cos A + cos 2A + cos 4A + cos 5A + cos 7A + cos 8A.

Recall that cos nA is the real part of einA . Then

Here ω is a complex cube root of unity.

L = {(a, b, c) + λ(p, q, r) | λ is a real number.}

(b) Find the equation of a plane obtained by rotating x + y = 0 about L by 45◦ .

Hence L can be written as

So the equation of the plane is

Choosing B = C and A+B+C = 0 we see that the coecient of x200 is (101)(102) 6=

Hence f (m) is either a square or is in A. Thus it is enough to assume that m ∈ A.

This clearly implies that f 2j (m) = (k + j)2 .

(a) Suppose a segment AC meets with opposite sides P Q and T S of a hexagon

P Q = T S =⇒ P A+AQ = T C +CS =⇒ AQ(k+1) = CS(k+1) =⇒ AQ = CS.

So 4AQO ∼ = 4CT O, so that OQ = OT . Also, ∠AOQ = ∠COT and ∠AOP =

180◦ = ∠OBA0 +∠OBQ = ∠OBQ+∠OA0 B = ∠OBQ+∠OAQ = 45◦ +β +45◦ +α,

so α + β = 0◦ , a contradiction since α + β = 60◦ .

1. Consider an equilateral triangle ABC with altitude 3 centimeters. A circle is inscribed

Find the sum of the areas of all these circles.

2. Consider the following function defined for all real numbers x

(a) How many complex solutions does this equation have?

p0 (x) = (x2 + 1)(x2 + 2) · · · (x2 + 1009)

1. Answer the following questions-

f ∗ (y) := max {yx − f (x)} ,

(a) Find all real solutions of the equation [6 marks]

Explain why your solutions

Answer : 2. Let the numbers be a, b respectively. Note a3 − b3 = 20 and

f (2n) = f (f (n)) and f (2n + 1) = f (2n) + 1.

(a) If f (0) = 0, find f (n) for every n. [2 marks]

i) no letter appears twice and,

Choosing B = C and A+B+C = 0 we see that the coecient of x200 is (101)(102) 6=