An alternative factorization

x 100 1 i > 0 ( x + a i ) ( m o d 101 ) \displaystyle { x }^{ 100 }-1\equiv \prod _{ i> 0 }^{ \quad }{ (x+{ a }_{ i }) } \pmod{101}

For positive integers a i , a_i, consider the congruence above. Observe that these are linear factors, so a standard factorization for integers will not work ( ( in fact you'll have a 5 0 th 50^\text{th} power in there ) . ). Determine the minimum sum of all a i . {a}_{i}.


The answer is 5050.

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.

2 solutions

Deeparaj Bhat
Dec 15, 2016

Note that p = 101 p = 101 is a prime. Hence, Z p \mathbb{Z}_p is a field.

Also, note that Fermat's theorem shows that 1 , 2 , , 100 1,2,\dots,100 are all roots of the following equation in Z p \mathbb{Z}_p : x 100 = 1 x^{100} = 1

As the degree of the polynomial is 100 100 , the above roots are the only roots.

As we're in a field, we conclude that the following holds in the field : x 100 1 = i = 1 100 ( x i ) = i = 1 100 ( x + i ) x^{100} - 1 = \prod_{i = 1}^{100} (x - i) = \prod_{i = 1}^{100} (x+i)

Also, this is the unique factorization (up to a permutation of the factors).

Hence, a i i ( m o d p ) i , ( ) i = 1 100 a i 5050 Equality holds when a i = i a_i \equiv i \pmod p \quad \forall i, (*)\\\implies \sum_{i = 1}^{100} a_i \geq 5050 \\\text{Equality holds when } a_i = i


Note: The equation ( ) (*) should be the following to be more accurate : { ( a i mod p ) : i N ; i 100 } = Z p { 0 } \{(a_i \text{ mod } p) : i \in \mathbb{N} ;\, i \leq 100\} = \mathbb{Z}_p - \{0\}

Dylan Pentland
Jun 30, 2015

Surprisingly, we can say that a i = i {a}_{i}=i . That is,

x 100 1 i > 0 100 ( x + i ) ( m o d 101 ) \displaystyle { x }^{ 100 }-1\equiv \prod _{ i> 0 }^{ 100 }{ (x+i) } \pmod{101}

Observe that x = 0 x=0 gives 1 -1 in modulo 101 101 . One can use Fermat's Little Theorem, that x p 1 1 ( m o d p ) {x}^{p-1}\equiv1 \pmod{p} to show that for x 0 x\neq0 we get 0 0 mod 101. This is essentially because if we take a nonzero x x then

{ x , 2 x , . . . ( p 1 ) x } = { 1 , 2 , 3... ( p 1 ) } \{x,2x,...(p-1)x \}=\{1,2,3...(p-1) \}

Taking the product of all elements in both sets and dividing by ( p 1 ) ! (p-1)! gives x p 1 1 ( m o d p ) {x}^{p-1}\equiv1\pmod{p} . So, we want a root for all 0 < x < 101 0<x<101 mod 101 which can clearly be done by setting a i = i {a}_{i}=i , due to each x x having a unique additive inverse mod 101. Using Wilson's theorem, ( p 1 ) ! 1 ( m o d p ) (p-1)!\equiv -1 \pmod{p} so x = 0 x=0 also works. Sum all the a i {a}_{i} :

i > 0 100 a i = ( 100 ) ( 101 ) 2 = 5050 \sum _{ i>0 }^{ 100 }{ { a }_{ i } } =\frac { (100)(101) }{ 2 } =5050

We cannot add to this factorization without increasing the sum of a i {a}_{i} : if the sum does not increase, any factors we add must be x x which invalidates the factorization at x = 0 x=0 . If we have ( x + k ) (x+k) then, well, you've just added k k to your minimum and it isn't very minimal anymore. :)

A few statements:

I don't think this proves Wilson's theorem; I think we have to use Wilson's theorem to prove that this is a correct solution. Nothing is preventing this factorization from not being 1 ( m o d 101 ) -1\pmod{101} when x 0 ( m o d 101 ) x\equiv 0\pmod{101} , which would make it incorrect.

Also, I think you need to make a brief statement about how any polynomial of degree > 100 >100 must have a higher sum.

Daniel Liu - 5 years, 11 months ago

Log in to reply

Yeah, the reasoning doesn't seem quite as solid as using Wilson's theorem so I've changed it. I also added a section on why you cannot add more terms. Thanks for feedback!

Dylan Pentland - 5 years, 11 months ago

Actually it does prove Wilson's theorem, if written correctly and using a theorem for polynomial congruences.It's actually the shortest proof of Wilson.

Bogdan Simeonov - 5 years, 11 months ago

This does indeed prove Wilson's theorem if one knows a thing or two about polynomial congruences.Basically, the following theorem is needed:(the modular equivalent of FTA)Every polynomial of degree n has at most n incongruent solutions mod p.Since we know that (x^{p-1}-1} has the roots 1,2,3...,p-1, we can now write the following congruence : x p 1 1 ( x + 1 ) ( x + 2 ) . . . ( x + p 1 ) ( m o d p ) x^{p-1}-1 \equiv (x+1)(x+2)...(x+p-1) (mod p) .Putting x=0 proves Wilson.

Bogdan Simeonov - 5 years, 11 months ago

Log in to reply

Ah, okay. I used the same reasoning about the roots before when I said it proved Wilson's but wasn't entirely sure about it (reasoning just seemed sketchy without knowing it was a theorem). Still, you use a theorem either way and the solution is shorter if you just use Wilson's :)

Dylan Pentland - 5 years, 11 months ago

You need to state that x x is an integer, and also state the degree of the polynomial.

Julian Poon - 4 years, 11 months ago

I'm not sure why I can't submit a solution — the correct answer is 1 1 .

For x = 100 x = 100 , we have x 100 1 0 ( m o d 101 ) x^{100} - 1 \equiv 0 \pmod{101} by Fermat's Little Theorem, since 101 101 is prime. Therefore,

x 100 1 x + a ( m o d 101 ) x^{100} - 1 \equiv x + a \pmod{101}

with x = 100 x = 100 and a = 1 a = 1 , satisfying all the conditions in the statement.

(This is my first comment, and it took me a while to figure out how to get the math notation right — it would be nice to have a link to the formatting guide nearby ;-))

Mika MF - 1 month ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...