User Name Remember Me? Password

 Math Events Math Events, Competitions, Meetups - Local, Regional, State, National, International

 June 18th, 2014, 07:54 PM #1 Senior Member   Joined: Sep 2012 From: British Columbia, Canada Posts: 764 Thanks: 53 Math Q&A Part 2 Let a divisor $d$ of a positive integer $n$ be called a perfect divisor if $\gcd (d,\frac{n}{d})=1$. Let $m$ be the sum of all perfect divisors of $2014!$. Find $m\,(\text{mod}\, 2014)$. June 18th, 2014, 09:09 PM #2 Math Team   Joined: Dec 2013 From: Colombia Posts: 7,683 Thanks: 2664 Math Focus: Mainly analysis and algebra 2014 = 2 x 19 x 53 Since all exponential of each prime factor is 1, every proper divisors of 2014 (i.e. not 2014 and 1) is a perfect divisor. So m = 2 + 1007 + 19 + 106 + 53 + 38 = 1225 = 1225 mod 2014 Thanks from eddybob123 June 18th, 2014, 09:24 PM   #3
Global Moderator

Joined: Nov 2006
From: UTC -5

Posts: 16,046
Thanks: 938

Math Focus: Number theory, computational mathematics, combinatorics, FOM, symbolic logic, TCS, algorithms
Quote:
 Originally Posted by eddybob123 Let a divisor $d$ of a positive integer $n$ be called a perfect divisor if $\gcd (d,\frac{n}{d})=1$. Let $m$ be the sum of all perfect divisors of $2014!$. Find $m\,(\text{mod}\, 2014)$.
The usual terminology is unitary divisor. The sum of the unitary divisors of 2014! is
$$\prod_{p\le2014}p^e+1$$
where the product is over primes and the exponent e is the p-adic valuation of 2014: $e=\left\lfloor\frac{2014}{p}\right\rfloor+\left\l floor\frac{2014}{p^2}\right\rfloor+\ldots$.

1861 is prime and divides 2014! exactly once (since 2014/2 < 1861 <= 2014) and so the product is divisible by 1862 = 2 * 7^2 * 19. Similarly, since 1907 is prime the product is divisible by 1908 = 2^2 * 3^2 * 53. Hence the product is divisible, at least, by 2 * 19 * 53 = 2014 and so the answer is 0 mod 2014.

For those interested in the nonmodular answer in its 5782-digit glory, here's a GP script:
Code:
val(n,p)=my(t);while(n\=p,t+=n);t
t=1;forprime(p=2,2014,t*=p^val(2014,p)+1);t June 19th, 2014, 04:26 AM #4 Math Team   Joined: Dec 2013 From: Colombia Posts: 7,683 Thanks: 2664 Math Focus: Mainly analysis and algebra Somehow I read that "!" as English punctuation! I thought it was a bit easy (although I woulld in the cold light of day add 1 and 2014 to my answer). Thanks from eddybob123 June 19th, 2014, 12:18 PM   #5
Senior Member

Joined: Sep 2012
From: British Columbia, Canada

Posts: 764
Thanks: 53

Quote:
 Originally Posted by CRGreathouse Hence the product is divisible, at least, by 2 * 19 * 53 = 2014 and so the answer is 0 mod 2014.
Not that I understood what you did... Anyways, it's your question! June 20th, 2014, 05:29 AM   #6
Global Moderator

Joined: Nov 2006
From: UTC -5

Posts: 16,046
Thanks: 938

Math Focus: Number theory, computational mathematics, combinatorics, FOM, symbolic logic, TCS, algorithms
Quote:
 Originally Posted by eddybob123 Not that I understood what you did...
Hmm, maybe I didn't explain it well enough.

Since you want unitary divisors only, the basic units are not the primes but the prime powers maximally dividing the number. So let N = q1 q2 ... qk where each qi is a prime power and no primes are reused. Then for each qi it can be included, or not, in a unitary divisor. Thus the number of unitary divisors is 2^k.

To get their sum you could examine each of these 2^k numbers, but you can do better by noticing that the unitary divisors are either divisible by qk, or not, and you can transform one group into the other by multiplying/dividing by qk itself. So if you knew the sum of the divisors of N/qk you can multiply it by 1 + qk to get the sum of the divisors of N. Induction gives you what you;d expect: you can just multiply qi + 1 for all prime powers qi || N (that is, all p^e | N where p^(e+1) does not divide N).

Quote:
 Originally Posted by eddybob123 Anyways, it's your question!
Hmm. Draw a regular hexagon and connect all pairs of points. How many triangles are there in total? June 20th, 2014, 05:40 AM #7 Math Team   Joined: Dec 2013 From: Colombia Posts: 7,683 Thanks: 2664 Math Focus: Mainly analysis and algebra Does this include joining two triangles to make another? Or indeed, joining shapes of any type to make triangles? Thanks from eddybob123 June 20th, 2014, 05:47 AM #8 Global Moderator   Joined: Nov 2006 From: UTC -5 Posts: 16,046 Thanks: 938 Math Focus: Number theory, computational mathematics, combinatorics, FOM, symbolic logic, TCS, algorithms Any triangle as long as all three sides are sides or diagonals of the regular hexagon. Thanks from eddybob123 June 20th, 2014, 05:50 AM #9 Math Team   Joined: Dec 2013 From: Colombia Posts: 7,683 Thanks: 2664 Math Focus: Mainly analysis and algebra But triangles with sides that are segments of a diagonal are not included? Thanks from eddybob123 June 20th, 2014, 06:05 AM   #10
Global Moderator

Joined: Nov 2006
From: UTC -5

Posts: 16,046
Thanks: 938

Math Focus: Number theory, computational mathematics, combinatorics, FOM, symbolic logic, TCS, algorithms
Quote:
 Originally Posted by v8archie But triangles with sides that are segments of a diagonal are not included?
Those are included, too. (I was looking at everything as a line; if you look at segments instead you should use those. It all comes to the same thing.) Tags math, part, qanda ,

,

,

,

### math q&a in bengali

Click on a term to search for related topics.
 Thread Tools Show Printable Version Email this Page Display Modes Linear Mode Switch to Hybrid Mode Switch to Threaded Mode Similar Threads Thread Thread Starter Forum Replies Last Post eddybob123 Math Events 32 December 17th, 2013 09:12 PM Hoempa Math Events 109 August 10th, 2013 09:46 AM Denis Math Events 190 July 22nd, 2013 12:21 AM Denis Math Events 64 July 6th, 2013 08:33 AM mathbalarka New Users 8 June 28th, 2013 11:14 AM

 Contact - Home - Forums - Cryptocurrency Forum - Top      