Inspired by Calvin Lin

Algebra Level 3

a ( a 1 ) ( a 2 ) = a a 2 3 a + 2 \Large a^{(a-1)^{(a-2)}} = a^{a^2-3a+2}

Find the sum of all positive integers a a that satisfy the equation above.


The answer is 3.

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

Otto Bretscher
Apr 26, 2015

We check a = 1 , 2 , 3 a=1,2,3 and find that only a = 3 a=\boxed{3} is a solution. (For a = 1 a=1 , the exponent on the LHS, 0 1 0^{-1} is undefined.)

If a 4 a\geq{4} then a ( a 1 ) ( a 2 ) a ( a 1 ) 2 > a ( a 1 ) ( a 2 ) = a a 2 3 a + 2 a^{(a-1)^{(a-2)}}\geq{a^{(a-1)^2}}>a^{(a-1)(a-2)}=a^{a^2-3a+2} .

Moderator note:

Great use of inequalities on exponents. Nicely done.

Lovely! A simple and straightforward approach.

Quince Pan - 6 years ago

Why can't we use integers a>3?

vojtech filipi - 2 years, 7 months ago

First, it is tempting to automatically think that a = 1 a = 1 is a solution. However, while the expression on the RHS of the equation becomes 1 0 = 1 , 1^{0} = 1, the expression on the LHS becomes 1 0 1 , \large 1^{0^{-1}}, which cannot be evaluated since 0 1 = 1 0 0^{-1} = \dfrac{1}{0} is indeterminate. I'm open to counter-arguments, though, since even if x x is indeterminate, 1 x 1^{x} might still be considered to have a solution of 1. 1.

Having dealt with, (for the time being), the case of a = 1 , a = 1, for positive integers a > 1 a \gt 1 we will require that the exponents be equal to one another, i.e.,

( a 1 ) ( a 2 ) = a 2 3 a + 2 = ( a 1 ) ( a 2 ) . (a - 1)^{(a - 2)} = a^{2} - 3a + 2 = (a - 1)(a - 2).

Now since ( a 1 ) 0 (a - 1) \ne 0 we can divide through by ( a 1 ) (a - 1) to end up with

( a 1 ) ( a 3 ) = a 2. (a - 1)^{(a - 3)} = a - 2.

Now for a positive integer, raised to an integer power, to yield a value strictly less than the original integer, we must have that power be non-positive. Now since a > 1 , a \gt 1, we can only ( a 3 ) 0 (a - 3) \le 0 when either a = 2 a = 2 or a = 3. a = 3.

For a = 2 a = 2 we have ( 2 1 ) 2 3 = 1 0 = 2 2. (2 - 1)^{2 - 3} = 1 \ne 0 = 2 - 2.

For a = 3 a = 3 we have ( 3 1 ) 3 3 = 1 = 3 2 , (3 - 1)^{3 - 3} = 1 = 3 - 2, and thus a = 3 a = 3 is the only possible solution. Checking this value in the original equation, the LHS becomes 3 2 1 = 9 , \large 3^{2^{1}} = 9, and the RHS becomes 3 ( 9 9 + 2 ) = 9 , \large 3^{(9 - 9 + 2)} = 9, thus confirming that a = 3 a = 3 satisfies the original equation.

Since a = 3 a = 3 is the only valid solution, the sum of all positive integer solutions is just 3 . \boxed{3}.

Moderator note:

Great analysis! What other values of a a satisfy the equation if the constraint a N a \in \mathbb N is relaxed, that is find all solutions such that the given equation is fulfilled.

You're searching for 1 0 \sqrt[0]{1} . The n th n^\text{th} root is the inverse of the x x n x \mapsto x^n . For n = 0 n=0 , the operation is not invertible because it's a constant function. So there's no such thing as a 0th root.

Pi Han Goh - 6 years, 1 month ago

Log in to reply

One could argue that the 0th root of 1 is (very) multivalued (as roots often are): It consists of all (nonzero) numbers.

There is no debating, though, that 1 0 1 1^{0^{-1}} is undefined. We are asked to evaluate the exponent, 0 1 0^{-1} , first, and that exponent is undefined.

Otto Bretscher - 6 years, 1 month ago

Log in to reply

if a=1 then lhs=1^0^-1=1(1^infinity) and rhs is 1.so a=1 is also a solution

arun kumaran - 5 years, 6 months ago

Log in to reply

@Arun Kumaran No, LHS is undefined

Otto Bretscher - 5 years, 6 months ago

O.k., great, that does it; a = 1 a = 1 is definitely not a solution. :)

Brian Charlesworth - 6 years, 1 month ago

As expected ... Nice!

Nihar Mahajan - 6 years, 1 month ago

Log in to reply

Thanks, Nihar. Great question. :)

Brian Charlesworth - 6 years, 1 month ago

What did you expect?

Swapnil Das - 6 years, 1 month ago

Log in to reply

I expected a good solution from Brian Sir.

Nihar Mahajan - 6 years, 1 month ago

Log in to reply

@Nihar Mahajan They are always the best in quality.

Swapnil Das - 6 years, 1 month ago

Log in to reply

@Swapnil Das I t o t a l l y a g r e e w i t h y o u . I~~\color{#D61F06}{ \large totally} ~~agree ~~with ~~you.

Niranjan Khanderia - 6 years, 1 month ago

To find the other value of a if the constraint is removed: I used the Desmos.com graphing calculator (Wonderful site: highly recommended. It also highlights key points of interest.... where two graphs interset, x- and y-intercepts, and local maxima and minima.)

First I put the LHS and RHS in as equations 1 & 2. Then I put in a third equation where I subtracted the RHS from the LHS and where the difference is zero, and the graph crosses the x-axis, it shows where they are equal. x^{\left(\left(x-1\right)^{\left(x-2\right)}\right)}-\left(x^{\left(x^2-3x+2\right)}\right)

x=3, y=9 is verified as a solution.

The other solution is x=3.354, y=47.33

What is striking about these graphs is the steepness of the slope of all three equations above x=3.6.

Sue Harris - 6 years, 1 month ago

Log in to reply

TI graphing calculators are are also good. I have been using it, and find all the above are available in it + a single answer programming apart from programming extra- which I have not used. This is no way to promote TI. It is only to say " those with TI can use single output programming facility through its ENTRY key."

Niranjan Khanderia - 6 years, 1 month ago

Here's another way to look at it.

We note the fact that f ( x ) = ( x 1 ) ( x 2 ) f(x)=(x-1)^{(x-2)} grows faster than g ( x ) = ( x 1 ) ( x 2 ) g(x)=(x-1)(x-2) after x = 3 x=3 . That is to say,

( x 1 ) ( x 2 ) o ( ( x 1 ) ( x 2 ) ) (x-1)(x-2)\in\mathcal{o}\left((x-1)^{(x-2)}\right)

This can be done easily either by graphing or using derivative tests. After that, we're simply left with checking the equation with x = 0 , 1 , 2 , 3 x=0,1,2,3 since the restriction x N x\in\Bbb{N} is imposed.

Note: o ( ) \mathcal{o}(\cdots) denotes Little-o notation .

Note-2: Take x = a x=a .

Prasun Biswas - 6 years, 1 month ago

I simply love this solution.

EKENE FRANKLIN - 2 years, 11 months ago

I also did the same.

Laxmi Narayan Bhandari Xth B - 8 months, 3 weeks ago

The question should be of level 3 or 4.

Swapnil Das - 6 years, 1 month ago

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...