Another Divisors Problem

Let D n D_n be the product of all positive divisors of a positive integer n . n. ( \big( For example, D 1 = 1 D_1 = 1 and D 4 = 1 × 2 × 4 = 8. ) D_4 = 1 \times 2 \times 4 = 8.\big)

What is the smallest positive integer n n for which D n D_n can be written as D n = p a × q b × r c × s d , D_n = p^a \times q^b \times r^c \times s^d, where p , q , r , s p, q, r, s are four distinct prime numbers and a , b , c , d a,b,c,d are four distinct positive integers?


Bonus 1: Can you solve this for prime numbers p p , q q , r r , s s , t t and integers a , b , c , d , e ? a,b,c,d,e?
Bonus 2: Can you generalize for p 1 , . . . , p n , p_{1}, ...,p_{n}, where all primes in the prime factorization are raised to different powers?


The answer is 75600.

This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try refreshing the page, (b) enabling javascript if it is disabled on your browser and, finally, (c) loading the non-javascript version of this page . We're sorry about the hassle.

10 solutions

Patrick Corn
Mar 5, 2018

Let d ( n ) d(n) be the number of divisors of n . n. Then D n = n d ( n ) / 2 . D_n = n^{d(n)/2}.

Proof: D n 2 = ( d n d ) ( d n d ) = ( d n d ) ( d n n d ) = d n n = n d ( n ) . \begin{aligned} D_n^2 &= \left( \prod_{d|n} d \right) \left( \prod_{d|n} d \right) \\ &= \left( \prod_{d|n} d \right) \left( \prod_{d|n} \frac{n}{d} \right) \\ &= \prod_{d|n} n = n^{d(n)}. \end{aligned}

For the conditions of the problem, D n D_n has the same prime divisors as n , n, so it must have four distinct prime divisors. The exponents in D n D_n are distinct if and only if the exponents in n n are distinct, so we're looking for the smallest number n n with four distinct prime divisors with four distinct exponents in the factorization of n . n. This is clearly 2 4 3 3 5 2 7 = 75600 . 2^4 3^3 5^2 7 = \fbox{75600}. The generalization to more prime divisors is immediately clear.

Very nice explanation. The same way I did it.

Santosh Tiwari - 3 years, 3 months ago

And with this explanation bonus 2 is trivial.

Pau Cantos - 3 years, 3 months ago

This is the same answer that I got. I then thought I misread the question and that caused me to put the answer in wrong.

alex schroeder - 3 years, 2 months ago

Oh,I thought that a,b,c,d,p,Q,r,s all are distinct

Mr. India - 3 years, 2 months ago

Log in to reply

same then i came to this solution 11^2 3^6 5^4*7^1 = 385914375

robin jonckheere - 3 years, 2 months ago

i juat want to mention there is a small problem when n is a square number. In such case, the way you write it does not fit the definition since d=sqrt(n)=n/d will be multiplied twice.

Alex mak - 3 years, 2 months ago

Log in to reply

The formula works for all n , n, and the proof still works. Try it!

Patrick Corn - 3 years, 2 months ago
Leonel Castillo
Mar 2, 2018

I believe there may be simpler solutions that rely less on the formal theory of arithmetic functions, but here I have found a solution to the general problem that relies on first studying the function D n D_n . In particular, finding out a general formula for D n D_n .

First, by definition D n = d n d D_n = \prod_{d | n } d . Let's for now consider the related arithmetic function, log D n = d n log d \log D_n = \sum_{d | n} \log d . The purpose of this is that the theory of divisor sums is well established so it is always better to relate products to sums. Let's now assume that n = a b n = ab with gcd ( a , b ) = 1 \gcd(a,b) = 1 . Then log D n = log D a b = d a b log d = d a , e b log d e = d a , e b log d + log e = d a , e b log d + d a , e b log e = τ ( b ) log D a + τ ( a ) log D b \log D_{n} = \log D_{ab} = \sum_{d | ab} \log d = \sum_{d | a , e | b} \log de = \sum_{d | a , e | b } \log d + \log e = \sum_{d | a , e | b } \log d + \sum_{d | a , e | b } \log e = \tau(b) \log D_a + \tau (a) \log D_b .

In this last step, τ \tau denotes the divisor counting function. This last step is justified because when we split the sum as such, the sums depend only on one of the parameters but we need to add it once for every time that same sum is added again because of the second parameter. With this computation, we finally deduce D a b = e τ ( b ) log D a + τ ( a ) log D b D_{ab} = e^{ \tau(b) \log D_a + \tau (a) \log D_b} and simplifying, this just says D a b = ( D a ) τ ( b ) ( D b ) τ ( a ) D_{ab} = (D_a)^{\tau(b)} (D_b) ^{ \tau(a)} . This is not quite as simple as multiplicativity, but this identity is interesting enough to inspire us to exploit it. However, we must first discover how we can apply it to the given problem so for now let's study the problem itself:

If D x D_x is going to have only the n n prime factors, necessarily x x must have just those exact prime divisors. Therefore we can say that our solution will be of the form i = 1 n p i q i \prod_{i=1}^n p_i ^{q_i} . In this product, all of the distinct primes are obviously relatively prime so this inspires us to apply the previous identity to compute D i = 1 n p i q i D_{\prod_{i=1}^n p_i ^{q_i}} in terms of the D p i q i D_{p_i ^{q_i}} for this, we will go further and prove the following identity (where all the numbers are relatively prime in pairs):

D a 1 a 2 . . . a n = i = 1 n D a i j = 1 , j i n τ ( a j ) D_{a_1 a_2 ... a_n} = \prod_{i=1}^{n} D_{a_i} ^{ \prod_{j=1, j \neq i}^n \tau(a_j)}

We proceed by induction. We already proved the case n = 2 n=2 so let's now assume it works until a given n n and with that let's prove it for n + 1 n+1 . We have that D a 1 a 2 . . . a n a n + 1 = D ( a 1 . . . a n ) a n + 1 = D a 1 . . . a n τ ( a n + 1 ) D a n + 1 τ ( a 1 . . . a n ) = D a n + 1 τ ( a 1 ) . . . τ ( a n ) i = 1 n D a i τ ( a n + 1 ) j = 1 , j i n τ ( a j ) = i = 1 n + 1 D a i j = 1 , j i n + 1 τ ( a j ) D_{a_1 a_2 ... a_n a_{n+1}} = D_{(a_1 ... a_n ) a_{n+1}} = D_{a_1 ... a_n} ^{ \tau (a_{n+1})} D_{a_{n+1}} ^{\tau( a_1 ... a_n)} = D_{a_{n+1}} ^{\tau(a_1)...\tau(a_n)} \prod_{i=1}^{n} D_{a_i} ^{ \tau(a_{n+1})\prod_{j=1, j \neq i}^n \tau(a_j)} = \prod_{i=1}^{n+1} D_{a_i} ^{ \prod_{j=1, j \neq i}^{n+1} \tau(a_j)} as desired.

Let's now apply this theorem for a i = p i q i a_i = p_i ^{q_i} which satisfy the condition of relatively prime in pairs because they are all distinct primes. To do this, let's first compute D p i q i = d p i q i d = 1 × p i × p i 2 × . . . × p i q i = p i q i ( q i + 1 ) 2 D_{p_i^{q_i}} = \prod_{d | p_i ^{q_i}} d = 1 \times p_i \times p_i ^2 \times ... \times p_i^{q_i} = p_i ^{ \frac{q_i (q_i + 1)}{2}} . And, as the known formula tells us, τ ( p i q i ) = q i + 1 \tau (p_i^{q_i}) = q_i + 1 . Mixing this together with the previous theorem we can finally compute:

D p 1 q 1 . . . p n q n = i = 1 n p i q i ( q 1 + 1 ) ( q 2 + 1 ) . . . ( q n + 1 ) 2 D_{p_1 ^{q_1} ... p_n ^{q_n}} = \prod_{i=1}^n p_i ^{\frac{q_i (q_1 + 1)(q_2 + 1)...(q_n + 1)}{2}}

Notice that now we have actually proven a complete product formula for the D x D_x in terms of the prime factorization of x x . This is always very useful to solve problems like this. Notice that in the exponent of the primes, the only thing that changes is the factor of q i q_i which means that if we want no exponents to be equal then we must have that for i j , q i q j i \neq j, q_i \neq q_j . This is the core condition we must follow. Now, to complete the problem lets minimize solutions:

First, notice that we can choose any primes we like so let's choose the smallest primes p 1 = 2 , p 2 = 3 , . . . p_1 = 2, p_2 = 3, ... until p n = p_n = the n n th prime. We can also choose any exponents, as long as they are all distinct. To minimize we can again choose the smallest exponents, 1 , 2 , . . . , n 1,2,...,n . We must now assign each exponent to a given prime. To minimize the expression we really should give the smallest exponents to the biggest primes, as the biggest primes dominate. Therefore the solution is p n × p n 1 2 × . . . × 2 n p_n \times p_{n-1}^2 \times ... \times 2^n .

For the case n = 4 n=4 this is nothing more than 7 × 5 2 × 3 3 × 2 4 = 75600 7 \times 5^2 \times 3^3 \times 2^4 = 75600 .

Pedro Cardoso
Mar 11, 2018

If n n has a prime factorization with k k primes, D n D_n will also have the exact same k primes (but raised to bigger powers, obviously), so we know n n has exactly 4 primes. Also, if there are 2 primes raised to the same power in it's factorization, they will also be raised to the same power in the divisor product, as the amount of times a given prime appears in the factorization of some number's divisors doesnt depend on what prime it is. Thus, we know all 4 primes in n n are raised to different powers. The smallest number that satisfies those proprieties is 2 4 3 3 5 2 7 2^{4}3^{3}5^{2}7 . However, there is a bit of hand-waving here, it kind of feels natural that since 7's exponent in n n is smaller than 5's exponent, and so on, it will also be like this in D n D_n , but I haven't proved why this is true.

You also have to show that distinct powers in n n produce distinct powers in D n D_n ; the symmetry argument only shows that distinct powers in n n are necessary but not sufficient. Or at least compute D 75600 D_{75600} and check the powers are distinct.

Matt McNabb - 3 years, 2 months ago

Log in to reply

You are correct, it kind of feels natural that, since n's exponent in 7 is smaller than the exponenet in 5, which is smaller then the one in 3, and so on, it will also be like this in D n D_n , but I havent proved that's true, I'll mention it in the solution.

Pedro Cardoso - 3 years, 2 months ago
Michael Wang
Mar 12, 2018

Just use intuition. Different prime numbers, so start small: 2 2 , 3 3 , 5 5 , and 7 7 . Distinct positive integers: 1 1 , 2 2 , 3 3 , 4 4 . Searching for smallest positive integer, so put high powers to low base: 2 4 3 3 5 2 7 = 16 27 25 7 = 75600 2^4*3^3*5^2*7=16*27*25*7=75600

Only one link is missing from this: why should the powers of the prime factors of n be distinct? We only know that they are distinct for Dn.

Laszlo Kocsis - 3 years, 3 months ago

Log in to reply

Apologies, but I don't understand your doubt. The question states that the powers need to be distinct, so we make them as distinct as possible while keeping the value the lowest.

Aryaman Singh - 3 years, 2 months ago

Log in to reply

It's not a doubt, but a missing link from the solution. While it's true, but it's not obvious that having different powers of prime factors of Dn means the powers of the prime factors would be different for n, too. Pedro Cardoso's solution above explains it in the second sentence.

Laszlo Kocsis - 3 years, 2 months ago

This is the most obvious way to solve it. I really don’t understand the other solutions.

BHINESH PATEL - 3 years, 2 months ago

Short and simple.

Brett Wiechmann - 3 years ago
Arjen Vreugdenhil
Mar 12, 2018

Write n = p α q β r γ s δ n = p^\alpha\cdot q^\beta\cdot r^\gamma\cdot s^\delta . Then σ ( n ) = ( α + 1 ) ( β + 1 ) ( γ + 1 ) ( δ + 1 ) \sigma(n) = (\alpha+1)(\beta+1)(\gamma+1)(\delta+1) is the number of divisors of n n , and a = 1 2 α σ ( n ) a = \tfrac12\alpha\cdot \sigma(n) , and so on.

The exponents a , , d a, \dots, d are independent of the choice of p , , s p, \dots, s . Therefore we choose the smallest possible prime numbers: p , q , r , s = 2 , 3 , 5 , 7 p, q, r, s = 2, 3, 5, 7 ; and the exponents different but as small as possible, pairing the greatest exponent with the smallest prime etc.: α , β , γ , δ = 4 , 3 , 2 , 1 \alpha,\beta,\gamma,\delta = 4, 3, 2, 1 . Thus the smallest solution is n = 2 4 3 3 5 2 7 1 = 75 600 , D n = 2 240 3 180 5 120 7 60 . n = 2^4\cdot 3^3\cdot 5^2\cdot 7^1 = \boxed{75\,600},\ \ \ D_n = 2^{240}\cdot 3^{180}\cdot 5^{120}\cdot 7^{60}.

In general, for k k terms, let p i p_i be the i i th smallest prime number, then n = i = 1 k p i k i + 1 , D n = n ( k + 1 ) ! / 2 . n = \prod_{i=1}^k p_i^{k-i+1},\ \ \ D_n = n^{(k+1)!/2}.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
# https://brilliant.org/weekly-problems/2018-03-12/advanced/?p=5

from sympy import primefactors

def factors(n):
    return set(sum([[i, n//i] for i in xrange(1, int(n**0.5)+1) if not n%i], []))

def prime_factors(num):
    k = num
    prime_factors = {}
    for i in primefactors(num):
        count = 0
        while num % i == 0:
            num = num/i
            count += 1
        prime_factors[i] = count

    result = sorted(prime_factors.items() , key=lambda t : t[0] , reverse=True)

    num = k 
    primes = []
    ints = []
    for k,v in result:
        primes.append(k)
        ints.append(v)
    return ints, primes, result
#-------------------------------------------------------------------------------

product = 1
numb_of_uniques = 4
i = 1
while i > 0:
    x = sorted(factors(i))
    for j in x:
        product *= j
    ints, primes, result = prime_factors(product)
    product = 1
    if len(list(set(ints))) == len(list(set(primes))) and len(primes) == numb_of_uniques:
        print 'Smallest n where D_n has %d unique primes and %d unique powers: %d  ' \
            %(numb_of_uniques, numb_of_uniques,i) ,
        count = 1
        for k,v in result:
            if count < len(result):
                print '(%d ** %d) *' % (k,v),
                count += 1
            else:
                print '(%d ** %d)' % (k,v)   

        break
    i += 1

1
2
3
4
5
6
7
Smallest n where D_n has 2 unique primes and 2 unique powers: 12          D_n = (3 ** 3) * (2 ** 6)

Smallest n where D_n has 3 unique primes and 3 unique powers: 360            D_n = (5 ** 12) * (3 ** 24) * (2 ** 36)

Smallest n where D_n has 4 unique primes and 4 unique powers: 75600            D_n = (7 ** 60) * (5 ** 120) * (3 ** 180) * (2 ** 240)

Smallest n where D_n has 5 unique primes and 5 unique powers: 174636000           D_n =  (11 ** 360) * (7 ** 720) * (5 ** 1080) * (3 ** 1440) * (2 ** 1800)

How do you add code blocks?

Daniel Xiang - 3 years, 2 months ago

Log in to reply

https://brilliant.org/math-formatting-guide/ Look near bottom of section called Markdown Code

Michael Fitzgerald - 3 years, 2 months ago

Log in to reply

Thanks mate :)

Daniel Xiang - 3 years, 2 months ago
Sakibul Haque
Mar 16, 2018

Just take the smallest 4 prime numbers and smallest four integers and arrange them in the way where the product is the smallest. In this case the smallest primes are 2,3,5,7 and integers are 1,2,3,4 We know that the product of larger numbers are larger such as 7^4 > 2^4. So we just arrange the primes with power where the largest primes power is the smallest integer. This way the product will be the smallest and the value will be the product of 2^4,3^3,5^2,7^1 which is equal to 75600

Yash Ghaghada
Mar 15, 2018

Okay from the given conditions, it is clear that the number n n is made up of 4 4 prime nos p , q , r , s p,q,r,s

Let the number n = p α q β r γ s δ n=p^{\alpha }q^{\beta }r^{\gamma }s^{\delta }

now D n D_{n} is the product of all factors of n n . Let us think about that

  • Now in multiplication of all factors, there are these many elements with base p p i.e [ ( p 0 . . . p 0 ) ( p 1 . . . p 1 ) . . . . . ( p α . . . p α ) ] [(p^{0}...p^{0})(p^{1}...p^{1}).....(p^{\alpha }...p^{\alpha })] and similarly for q , r q,r and s s .

  • Now note that p 0 , p 1 , . . . , p α p^{0}, p^{1}, ... , p^{ \alpha} occurs ( β + 1 ) ( γ + 1 ) ( δ + 1 ) (\beta +1)\cdot (\gamma +1)\cdot (\delta +1) times (by counting the remaining possibilities)

  • So net power of p p in D n D_{n} is ( β + 1 ) ( γ + 1 ) ( δ + 1 ) × i = 0 α i (\beta +1)\cdot (\gamma +1)\cdot (\delta +1)\times\sum_{i=0}^{\alpha }i which gives ( α ) ( α + 1 ) ( β + 1 ) ( γ + 1 ) ( δ + 1 ) 2 \frac{(\alpha )\cdot (\alpha +1)\cdot(\beta +1)\cdot (\gamma +1)\cdot (\delta +1)}{2}
  • Let S = ( α + 1 ) ( β + 1 ) ( γ + 1 ) ( δ + 1 ) 2 S=\frac{(\alpha +1)\cdot(\beta +1)\cdot (\gamma +1)\cdot (\delta +1)}{2}
  • Doing similar things for q , r q,r and s s

With this idea D n = p α S q β S r γ S s δ S D_{n}=p^{\alpha S}q^{\beta S}r^{\gamma S}s^{\delta S}


now comparing, a = α S a=\alpha S , b = β S b=\beta S , c = γ S c=\gamma S , d = δ S d=\delta S . Note that they are all different by the terms α , β , γ , δ \alpha, \beta, \gamma, \delta . So for a , b , c a,b,c and d d to be distinct α , β , γ , δ \alpha, \beta, \gamma, \delta should be distinct

so now to create minimum n n , primes should be 2 , 3 , 5 , 7 2,3,5,7 and correspondingly our n n becomes 2 4 3 3 5 2 7 1 2^{4}3^{3}5^{2}7^{1} which is equal to 75600 \boxed{75600}

Similarly we can generalize for more primes

Daniel Xiang
Mar 14, 2018

I wrote a mathematica programme to find out n n

For[n = 2*3*5*7, n < 100000, n++, 
    list = FactorInteger[Product[k, {k, Divisors[n]}]];
    If[Length[list] == 4, 
        a = Indexed[Indexed[list, 1], 2];
        b = Indexed[Indexed[list, 2], 2];
        c = Indexed[Indexed[list, 3], 2];
        d = Indexed[Indexed[list, 4], 2];
        If[a != b && a != c && a != d && b != c && b != d && c != d, Print[n]];
    ];
];

which outputs 75600 75600

or you could run this mathematica code...

Select[Range@100000,Length@Union[Last/@FactorInteger[Times@@Divisors@#]]==4&]

and get the same result but brute force is tough for n=5 and higher...

Giorgos K. - 3 years, 2 months ago

Haha, I guess that works as well.

Aryaman Singh - 3 years, 2 months ago
Lucas Victorino
Mar 13, 2018

Make your friends give you the answer it works all the time LOL

If you have any friends

Xoris Onoma - 3 years, 3 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...