Set Average

A set S S contains some number of distinct integers, the smallest of which is 0 and the largest is 2015. What is the smallest possible average of the elements of S S to the nearest integer?


The answer is 62.

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

Chew-Seong Cheong
Apr 29, 2017

Relevant wiki: Extremal Principle - Problem Solving

Let S = { a k } 0 N S = \{a_k\}_0^N , where a k Z a_k \in \mathbb Z , a 0 = 0 a_0 = 0 is the minimum and a N = 2015 a_N=2015 is the maximum. Then the average of elements of S S is given by a k = k = 1 N a k N \overline{a_k} = \dfrac { \sum_{k=1}^N a_k}N . Since two elements 0 and 2015 are already fixed, we can find the minimum a k \overline{a_k} by including some n n smallest elements from 1, 2, 3, ... to then take the average by dividing the sum by n + 2 n+2 . Then a k \overline{a_k} is given by:

a k = k = 1 n k + 2015 n + 2 = 1 2 n ( n + 1 ) + 2015 n + 2 = 1 2 ( n ( n + 1 ) + 4030 n + 2 ) = 1 2 ( n ( n + 2 ) n 2 + 4032 n + 2 ) = 1 2 ( n 1 + 4032 n + 2 ) = 1 2 ( n + 2 + 4032 n + 2 3 ) By AM-GM inequality: n + 2 + 4032 n + 2 2 4032 61.99803147 Equality occurs when: n 61.5 \begin{aligned} \overline{a_k} & = \frac {\sum_{k=1}^n k + 2015}{n+2} \\ & = \frac {\frac 12n(n+1)+2015}{n+2} \\ & = \frac 12 \left(\frac {n(n+1)+4030}{n+2}\right) \\ & = \frac 12 \left(\frac {n(n+2)-n-2+4032}{n+2}\right) \\ & = \frac 12 \left(n-1+\frac {4032}{n+2}\right) \\ & = \frac 12 \left({\color{#3D99F6}n+2+\frac {4032}{n+2}}-3\right) & \small \color{#3D99F6} \text{By AM-GM inequality: } n+2+\frac {4032}{n+2} \ge 2 \sqrt{4032} \\ & \ge 61.99803147 & \small \color{#3D99F6} \text{Equality occurs when: }n \approx 61.5 \end{aligned}

Upon checking, we find that a k m i n = 62 \overline{a_k}_{min} = \boxed{62} , when n = 61 , 62 n=61, 62 .

When you write { a k } 0 2015 \{a_k\}_0^{2015} , usually that means { a 0 , a 1 , a 2 , , a 2015 } \{a_0, a_1, a_2, \ldots, a_{2015}\} . It's unusual to see it to mean the minimum is 0 and the maximum is 2015. Maybe phrase it better?

Ivan Koswara - 4 years, 1 month ago

Log in to reply

You are right. I have changed the phrase.

Chew-Seong Cheong - 4 years, 1 month ago

Log in to reply

Well, now you imply that S S has 2016 elements ( a 0 , a 1 , a 2 , , a 2015 a_0, a_1, a_2, \ldots, a_{2015} ), while it's certainly not the case.

Ivan Koswara - 4 years, 1 month ago

Log in to reply

@Ivan Koswara Sorry, I didn't save it properly.

Chew-Seong Cheong - 4 years, 1 month ago

Log in to reply

@Chew-Seong Cheong No, when you write " a 0 = 0 a_0 = 0 is the minimum, a 2015 = 2015 a_{2015} = 2015 is the maximum", then it implies your elements are a 0 , a 1 , a 2 , , a 2015 a_0, a_1, a_2, \ldots, a_{2015} . You should have written " a 1 = 0 a_1 = 0 is the minimum, a N = 2015 a_N = 2015 is the maximum".

Ivan Koswara - 4 years, 1 month ago

Log in to reply

@Ivan Koswara Sorry, I swear I have changed it and it doesn't appear.

Chew-Seong Cheong - 4 years, 1 month ago

Why was the question changed to give the answer 'to the nearest integer'? It's not necessary, and could make people think that it's a calc question...

Miles Koumouris - 4 years ago

Log in to reply

When I did the question, it was already "to the nearest integer". Was the original not so?

Regardless, one reason why it's "to the nearest integer" is because the answer format is an integer; not saying so can make people to think "okay, the answer is an integer, so I can make this implication that would otherwise be illegal if I didn't know that".

Ivan Koswara - 4 years ago

Log in to reply

The original was not so. Some people get put off by questions that require computers to solve, plus I was also sort of trying to make a subtle point that it can be done by hand. Besides, if I were trying to give information that the answer had to be an integer, I would mention that in the problem rather than letting people deduce information from the input format.

Miles Koumouris - 3 years, 2 months ago

The trick is to add the smallest possible integers (i.e. 1, 2, 3, etc.) as long as the number added is less than the current average.

If we add the integers up to n n , we end up with n + 2 n + 2 elements in the set, with a total of 1 + 2 + + n + 2015 = 1 2 n ( n + 1 ) + 2015. 1 + 2 + \cdots + n + 2015 = \tfrac12n(n+1)+2015. One way to approach this problem is by solving the quadratic equation average = 1 2 n ( n + 1 ) + 2015 n + 2 = n , \text{average} = \frac{\tfrac12n(n+1)+2015}{n+2} = n, but I will pursue a method of successive approximation.

If we take n = 0 n = 0 (so that the set only contains two integers) the average is obviously equal to 2015 / 2 = 1007 1 2 2015/2 = 1007\tfrac12 . Adding all integers up to n = 1007 n = 1007 gives the next approximation: an average of 509 543 / 1010 = 504.998 509\,543/1010 = 504.998 . Therefore we try the next approximation with n = 505 n = 505 , and so on. The table shows the successive steps:

n sum average 0 2017 1008.50 1007 509 543 504.99 504 129 275 255.48 255 34 655 134.84 134 11 060 81.32 81 5 336 64.29 64 4 095 62.05 62 3 095 62.00 \begin{array}{r|r|r} n & \text{sum} & \text{average} \\ \hline 0 & 2017 & 1008.50 \\ 1007 & 509\,543 & 504.99 \\ 504 & 129\,275 & 255.48 \\ 255 & 34\,655 & 134.84 \\ 134 & 11\,060 & 81.32 \\ 81 & 5\,336 & 64.29 \\ 64 & 4\,095 & 62.05 \\ 62 & 3\,095 & 62.00 \\ \hline \end{array}

Thus the average is exactly 62 \boxed{62} , for the set { 0 , 1 , 2 , , 61 , 62 , 2015 } \{0,1,2,\dots,61,62,2015\} . Of course, removing 62 from the set would not affect the average.

0 pending reports

×

Problem Loading...

Note Loading...

Set Loading...