Gridlock

Use 25 distinct positive integers to fill up the grid, such that the integers in any two connected squares have a greatest common divisor that is greater than 1.

What is the smallest possible value of the largest positive integer used?


Variation on a Theme .


The answer is 33.

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.

1 solution

To establish an upper bound, note that we could use the first 25 positive even integers in any configuration, in which case the gcd of adjacent numbers will always exceed 1. This gives us an upper bound of 50 50 for the smallest maximum. Also, we cannot include 1 1 , as the gcd of it and any other positive integer will be 1. Thus at best we would be using the integers 2 2 through 26 26 .

However, consider what happens with primes 17 \ge 17 . Each square is adjacent to at least two other squares, so any number included must have two other numbers with which it shares a non-unit gcd. With 17 17 , we would have to include 34 34 and 51 51 at best, but as we know that 50 50 is an upper bound we know that 17 17 cannot be part of an optimal solution. The same goes for 19 19 and 23 23 . So at this point the best we could do is find a solution that involves the first 25 positive integers greater than 1 and excluding 17 , 19 17, 19 and 23 23 . This would raise our largest integer from 26 26 to 29 29 , but then 29 29 would have the same issue as 17 17 , so we would drop 29 29 and add 30 30 .

But now we have to deal with 11 11 and 13 13 . If we include 11 11 , we would have to include 33 33 , (with 22 22 already in play), and if we include 13 13 we would have to include 39 39 , (with 26 26 already in play). If we dropped 11 11 and 13 13 then the next available integers to add to the list would be 32 32 and 33 33 , (since 31 31 would have the same issues as 17 17 ). This is one possibility, but if we keep 11 11 and only drop 13 13 then we would have to add 33 33 as the 25th integer for reasons already discussed. So either way the best we can do is to work with a group of 25 integers the maximum of which is 33 33 , with the latitude to drop one otherwise suitable number less than 33 33 from the grid. So if we can create a suitable grid with this group of 25 integers, (with one degree of flexibility), then the smallest maximum would be 33 \boxed{33} . Here is such a grid, (with 27 27 chosen to be excuded):

7 21 18 15 3 \Large 7 \space \space 21 \space \space 18 \space \space 15 \space \space 3

14 28 2 12 9 \Large 14 \space 28 \space \space 2 \space \space 12 \space \space 9

4 8 16 26 24 \Large 4 \space \space \space 8 \space \space 16 \space \space 26 \space \space 24

10 20 32 6 22 \Large 10 \space 20 \space \space 32 \space \space 6 \space \space 22

5 25 30 33 11 \Large 5 \space \space 25 \space \space 30 \space \space 33 \space \space 11

I think you meant Greatest Common Divisor and not G.C. denominator

Am I mistaken?

Rohith M.Athreya - 4 years, 2 months ago

Log in to reply

You are right. Thanks for catching that. :) And congrats on being the only person to solve the problem so far.

Brian Charlesworth - 4 years, 2 months ago

Log in to reply

Firstly,I am sorry for a v e r y v e r y \displaystyle \large very^{very} late reply. Somehow,this notification had escaped me.

Thank you for your compliment!

I followed the same strain of thought which you have outlined in your solution,but have been trying to extend the results to matrices of higher dimensions.I must add that I haven't exactly gotten anywhere close.

If possible,can you please suggest an opening to the same?

Rohith M.Athreya - 4 years, 1 month ago

i started going down the path of Brian, realized I had to start with 2, recognized 3 was not going to be a fit, then extrapolated that all prime numbers would likewise not be a fit, and then simply added the count of prime numbers to 25, count(1, 3, 5, 7, 11, 13, 17, 23)=8 and added that to 25 and arrived at 33, all in less than a minute. Was this a coincidental lucky guess?

John Wicker - 4 years, 1 month ago

Log in to reply

Well, perhaps a bit lucky since you missed 19 in your tally of odd primes, but had 1 in your list, (which is not a prime), which had the effect of "making up" for the absence of 19 in your list. As clearly 1 cannot be used in the grid, your approach would exclude 1 and the 8 primes less than 25 for a total of 9 exclusions, which would lead to 25 + 9 = 34 as your maximum, (by your approach). Also, in excluding 1 and the primes less than 25 you would have only 16 positive integers 25 \le 25 remaining, which you would have to make up for by using the next 9 integers greater than 25, two of which are prime. So I think in the end, your arrival at 33 was a fortunate guess. :)

Note that my sample solution grid does include 3,5,7 and 11, so you don't need to rule out odd primes right from the start.

Brian Charlesworth - 4 years, 1 month ago

@Brian Charlesworth: I think that "... consider what happens with primes >= 11..." in your solution maybe okay!

Nguyen Nam - 4 years, 1 month ago

will the 6 × 6 6 \times 6 grid has a solution equal to 3 × 13 3 \times 13 and so on?

Mehdi K. - 4 years, 1 month ago

Log in to reply

With the 6 x 6 grid I find that we can introduce 13 13 into the list of primes used, leading to a smallest maximum of 45 45 . With the 7 x 7 grid we can introduce 17 17 and 19 19 as well, leading to a smallest maximum of 58 58 . With the 8 x 8 grid we can introduce 23 23 , leading to a smallest maximum of 77 77 . With the 9 x 9 grid we can introduce 29 29 and 31 31 , leading to a smallest maximum of 85 85 . There won't be a general formula because the prime gaps are uneven, but the general idea is to introduce successive primes into the list to try and reduce the maximum number used. The grids themselves actually become easier to form as they get larger as there are more options available.

@Rohith M.Athreya I just noticed your comment below. I'm sorry for not replying earlier, but my response to Mehdi should address your question to some degree.

Brian Charlesworth - 4 years, 1 month ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...