A Number Theoretical Nod To 2015

Suppose the product of n n consecutive, positive four-digit integers is divisible by 201 5 2 2015^{2} . What is the least possible value of n n ?


The answer is 11.

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.

3 solutions

We first note that 2015 = 5 × 13 × 31 2015 = 5 \times 13 \times 31 , and so 201 5 2 = 5 2 × 1 3 2 × 3 1 2 2015^{2} = 5^{2} \times 13^{2} \times 31^{2} . Now as we want the product of the consecutive integers to have at least two factors of 31 31 , we can either have two different integers in the sequence that are divisible by 31 31 or one that is divisible by 3 1 2 31^{2} . To have two divisible by 31 31 means that there would be at least 32 32 integers in the sequence, so we'll try instead to find a shorter sequence which includes one integer divisible by 3 1 2 = 961 31^{2} = 961 .

The four-digit integers divisible by 961 961 are 1922 , 2883 , 3844 , 4805 , 5766 , 6727 , 768 , 8649 1922, 2883, 3844, 4805, 5766, 6727, 768, 8649 and 9610 9610 . Now we also require that the product have at least two factors of 13 13 , so we can have two integers each divisible by 13 13 or one divisible by 1 3 2 = 169 13^{2} = 169 . In the former case we would then require a sequence of at least 14 14 integers, so let us first see if we can find an integer, divisible by 169 169 , with an absolute difference from one of the above list of multiples of 961 961 of less than 14 14 . Calculating these multiples modulo 169 169 , we find that the least absolute difference is 10 10 , as 2883 10 ( m o d 169 ) 2883 \equiv 10 \pmod{169} . This gives us the sequence 2873 2873 through 2883 2883 , the product of which also has 3 3 factors of 5 5 , (two from 2875 2875 and one from 2880 2880 ). Thus the smallest value of n n is 2873 2773 + 1 = 11 2873 - 2773 + 1 = \boxed{11} .

Cool problem! Yeah likewise I basically tried to find the minimum of 3 1 2 x 1 3 2 y |31^2x-13^2y| with positive integers x , y x,y , and it suffices to run 3 1 2 x 1 3 2 \displaystyle \frac {31^2x}{13^2} through the calculator for x = 2 , 3 , . . . , 10 x=2,3,...,10 and check if their decimal parts are close to 0 0 or 1 1 . e.g. 3 1 2 3 1 3 2 = 17.059... \frac {31^2\cdot 3}{13^2}=17.059...

Xuming Liang - 5 years, 4 months ago

Log in to reply

Thanks! Yes, there is a bit of number-crunching, but it only takes a minute or two. I went with 201 5 2 2015^{2} rather than 201 6 2 2016^{2} because I was sure of the answer for the former but not the latter. For 201 6 2 2016^{2} I think that the minimum for n n is 5 5 , with the product being 6908 6909 6910 6911 6912 6908*6909*6910*6911*6912 , but I would need to do more number-crunching to be absolutely sure.

Brian Charlesworth - 5 years, 4 months ago

Log in to reply

5 5 for 201 6 2 2016^2 is not hard to verify. It suffices to look at: 512 x 49 y < 5 |512x-49y|<5 for x = 2 , 3 , . . . , 19 x=2,3,...,19 . A quick decimal check (which can be further facilitated by considering 22 x ( m o d 49 ) 22x \pmod {49} ) gives us x = 9 , 18 x=9,18 , neither of which works because of that other divisibility: 3 4 3^4 .

Xuming Liang - 5 years, 4 months ago

Log in to reply

@Xuming Liang O.k., great. Thanks for the verification. :)

Brian Charlesworth - 5 years, 4 months ago
Arjen Vreugdenhil
Jan 28, 2016

The product must contain each of the prime factors 5, 13, 31 at least twice.

If n 31 n \leq 31 , then one of the numbers should be a multiple of 3 1 2 31^2 . Likewise, if n 13 n \leq 13 , then one of them should be a multiple of 1 3 2 13^2 . If n 10 n \geq 10 , we are guaranteed to have at least two factors 5, so we won't worry about that factor yet.

So I need multiples of 1 3 2 13^2 and 3 1 2 31^2 that are relatively close together. A good place to start is the ratio 3 1 2 1 3 2 5.686. \frac{31^2}{13^2} \approx 5.686. We like to approximate this ratio as a fraction a / b a/b , where a a and b b are sufficiently small; then 3 1 2 b 1 3 2 a 31^2\:b \approx 13^2\:a define the sequence of numbers that we look for. The requirement of four digits limits us to b 10 b \leq 10 and a 59 a \leq 59 .

The only candidates for approximating 5.686 5.686\dots with a fraction of denominator no greater than 10 are 5 1 2 = 5.500 5\tfrac12 = 5.500 (poor), 5 2 3 5.667 5\frac 23 \approx 5.667 ; 5 5 7 5.714 5\tfrac 57\approx 5.714 ; 5 7 10 = 5.700 5\tfrac 7{10} = 5.700 . The lower denominators will result into smaller values for n n .

  • 5 1 2 a = 11 , b = 2 5\tfrac12\ \to\ a = 11,\ b = 2 gives 3 1 2 b = 1922 31^2\ b = 1922 and 1 3 2 a = 1859 13^2\ a = 1859 , which are rather far apart. We can do better!

  • 5 2 3 a = 17 , b = 3 5\tfrac23\ \to\ a = 17,\ b = 3 gives 2883 2883 and 2873 2873 , so that n = 11 n = 11 .

  • 5 5 7 a = 40 , b = 7 5\tfrac57\ \to\ a = 40, b = 7 gives 6727 6727 and 6760 6760 , so that n = 34 n = 34 .

  • 5 7 10 a = 57 , b = 10 5\tfrac7{10}\ \to\ a = 57, b = 10 gives 9633 9633 and 9610 9610 , so that n = 24 n = 24 .

Our best option, then, is the second one, with n = 11 n = \boxed{11} . In any sequence of 10 or more successive numbers, there are at least two multiples of 5, so that condition is fulfilled.

Nice solution, Arjen. I like your fractional approximation method; it makes for a more efficient general approach than mine.

Brian Charlesworth - 5 years, 4 months ago

Hi Arjen,

I like your approach using approximations of 3 1 2 1 3 2 \frac{31^2}{13^2} . In your final statement you write that in any sequence of 10 or more successive numbers, there are at least two multiples of 5. This is indeed correct. However, earlier you remark "If n 10 n \leq 10 , we are guaranteed to have at least two factors 5". This is only true when n 10 n \geq 10 , so probably just a typo, right?

Kind regards, Patrick

Patrick Heebels - 5 years, 4 months ago
Soumava Pal
Feb 1, 2016

@Brian Charlesworth Nice one this problem... seems easy but is a great application of the greedy algorithm. We just need to check the shortest distance between multiples of 31^2 and 13^2 and I believe the answer would be less than 11 if there was no condition on the number of digits.

If you put no constraint at all on the size of the integers, you could have n = 1 n=1 , picking the single number 201 5 2 2015^2 .

If we allowed 5 5 -digit numbers, then since 15375 = 5 2 × 615 15375 = 5^2 \times 615 , 15376 = 3 1 2 × 16 15376 = 31^2 \times 16 and 15379 = 1 3 2 × 91 15379 = 13^2 \times 91 , we could achieve n = 5 n=5 . That is the best possible for 5 5 -digit numbers.

Mark Hennings - 5 years, 4 months ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...