Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Let 3,8,17,32,57 . . . . . be a pattern.How do we find the nth number?My brains are completely jammed,I am tired.I do not even recognize the pattern.I calculated a few ways,but all I want is a little hint,not the whole solution.

share|cite|improve this question
Any finite sequence can be extended in literally any way you want. There is no answer for what the sequence "should" be. – Zev Chonoles May 16 '13 at 17:46
@Zev Why nitpick? Yes, that's true, but I suspect (and this isn't a terribly difficult matter to ascertain) that the OP is looking for a satisfactory way to define the terms, assuming there is a pattern... – amWhy May 16 '13 at 17:48
There you go,I edited it,happy? – rahul May 16 '13 at 17:50
up vote 0 down vote accepted

Note: I just added a derivation of an explicit formula for the terms as the OP requested.

From 3,8,17,32,57, if the term is n, the next term is 2n+k where k = 2, 1, -2, -7.

The differences of k are -1, -3, -5.

If we assume that the next difference is -7, the next k is -7-7=-14 and the next term is 2*57-14 = 100.

To get a formula from this, since the sum of the odd numbers are the squares, the values of k are $2-m^2$ starting with $m = 0$.

Letting $s(0) = 3$, $s(n+1) = 2s(n)+2-n^2$. This is made explicit below.

Check: $s(1) = 2*3+2-0 = 8$, $s(2) = 2*8+2-1 = 17$, $s(3) = 2*17+2-4 = 32$, $s(4) = 2*32+2-9 = 57$.

To get an explicit form for $s(0) = 3$, $s(n+1) = 2s(n)+2-n^2$ let $s(n) = 2^n t(n)$. Then $2^{n+1}t(n+1) = 2^{n+1}t(n)+2-n^2$, or $t(n+1) = t(n) + 1/2^{n}-n^2/2^n$, so $t(n+1) - t(n) = 1/2^{n}-n^2/2^n$.

Summing this, $t(m)-t(0) = \sum_{n=0}^{m-1} (t(n+1) - t(n)) = \sum_{n=0}^{m-1} (1/2^{n}-n^2/2^n) = 1-1/2^m - \sum_{n=0}^{m-1}n^2/2^n $.

To evaluate the last sum, let $f(x) = \sum_{n=0}^{m-1} x^n = (1-x^m)/(1-x) $. Then $x f'(x) = \sum_{n=0}^{m-1} nx^n$ and $x (x f'(x))' = \sum_{n=0}^{m-1} n^2 x^n$.

We want $x (x f'(x))'$ at $x = 1/2$.

Note that $x (x f'(x))' = x (f'(x) + x f''(x)) = x f'(x) + x^2 f''(x) $.

I'm feeling lazy, so I won't bother getting $f'$ and $f''$.

share|cite|improve this answer

If you want the smallest degree polynomial that gives these numbers (which is only one of the infinitely many possible explanations for them), we can use the calculus of finite differences: $$\fbox{3}\quad 8\quad 17\quad 32\quad 57 \\ \fbox{5}\quad 9\quad 15\quad 25\\ \fbox{4}\quad 6\quad 10\\ \fbox{2} \quad 4\\ \fbox{2}$$ Define $$\begin{align*} f(n)&=3\binom{n}{0} +5\binom{n}{1}+4\binom{n}{2}+2\binom{n}{3}+2\binom{n}{4}\\\\ &=3 + \frac{19 n}{6} + \frac{23 n^2}{12} - \frac{n^3}{6} + \frac{n^4}{12} \end{align*}$$ Then $$f(0)=3,\quad f(1)=8,\quad f(2)=17,\quad f(3)=32,\quad f(4)=57$$

share|cite|improve this answer
Ah,I was counting 32=17+16.That is why I could find no pattern. – rahul May 16 '13 at 17:56

Hint: One solution can be obtained by taking differences-of-differences.

share|cite|improve this answer
Already did.Did not seem to make much sense,but I will concentrate a bit more. – rahul May 16 '13 at 17:51
The next number in the differences-of-differences can be chosen to be 16 to complete one possible pattern. – response May 16 '13 at 17:54

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.