What Is the Pattern in This Number Sequence?

  • Context: High School 
  • Thread starter Thread starter Alexx1
  • Start date Start date
Join the discussion
Ask a follow-up here, or get your own question answered by working scientists, mathematicians and engineers — people, not an autocomplete.
Real named experts · corrections over time · the nuance an AI answer skips
13 replies · 5K views
Alexx1
Messages
86
Reaction score
0
Can someone help me with this sequence?

1 4 1 16 1 36

If I only look at the odd numbers it's: n^2

But I don't know how they get to '1'..
 
Mathematics news on Phys.org


My answer would be to construct a step function for the y power term in x^y. The step function is a continuous representation of the integral of kronecker/dirac delta. The condition of the step is 'evenness' or 'oddness'.

Eg. If your series were,
1 8 1 64 1 216..

then my solution would be
[tex] n^{g};[/tex]
[tex] g=\frac{2}{1+ exp(P*(3*binx_n -1)};[/tex]
where binx_n is the last (right-most) digit in the binary representation of n; and P is a very large number (say 1e9). So in your case, the series will be: (read from left to right.. solution from MATLAB for 100 term series)

1 4 1 16 1 36 1 64 1 100 1 144 1 196 1 256 1 324 1 400 1 484 1 576 1 676 1 784 1 900 1 1024 1 1156 1 1296 1 1444 1 1600 1 1764 1 1936 1 2116 1 2304 1 2500 1 2704 1 2916 1 3136 1 3364 1 3600 1 3844 1 4096 1 4356 1 4624 1 4900 1 5184 1 5476 1 5776 1 6084 1 6400 1 6724 1 7056 1 7396 1 7744 1 8100 1 8464 1 8836 1 9216 1 9604 1 10000
...

So the big question here becomes, is the representation actually continuous (for it to be a nice analytical solution)? The presence of the binary makes that question hard to answer. Obviously, using the binary is an overkill. But there's a ton of ways to flip the exponential's sign given a number is odd or even. Any answer anyone else comes up with (your instructor, maybe), will be reducible to the given form.

Delightful little question. Hope this shows you how to consider such problems.

Edit: Note-My expression is different from what you need! If you want to test your understanding, then try to express this series with a continuous, differentiable expression:
1 4 1 64 1 36 1 512 1 100 1 1728 1 …
 
Last edited:


Alexx1 said:
Can someone help me with this sequence?

1 4 1 16 1 36

If I only look at the odd numbers it's: n^2

But I don't know how they get to '1'..
So you are starting with [itex]a_0[/itex]? There is nothing at all wrong with:
[itex]a_n= 1[/itex] if n is even, [itex](n+1)^2[/itex] if n is odd.

If you insist upon a single formula,
[tex]a_n= \frac{1- (-1)^n}{2}+ \frac{1+ (-1)^n}{2} (n+1)^2[/tex]

When n is even [itex](-1)^n= 1[/itex] so [itex]\frac{1- (-1)^n}{2}= 0/2= 0[/itex] and[itex]\frac{1+ (-1)^n}{2}= 2/2= 1[/itex]. Then [itex]a_n= 0+ (n+1)^2= (n+1)^2[/itex].

When n is odd [itex](-1)^n= -1[/itex] so [itex]\frac{1- (-1)^n}{2}= 2/2= 1[/itex] and [itex]\frac{1- (-1)^n}{2}= 0[/itex]. Then [itex]a_n= 1+ 0(n+1)^2= 1[/itex].
 
Last edited by a moderator:


HallsofIvy said:
So you are starting with [itex]a_0[/itex]? There is nothing at all wrong with:
[math]a_n= 1[/math] if n is even, [itex](n+1)^2[/itex] if n is odd.

If you insist upon a single formula,
[tex]a_n= \frac{1- (-1)^n}{2}+ \frac{1+ (-1)^n}{2} (n+1)^2[/tex]

When n is even [itex](-1)^n= 1[/itex] so [itex]\frac{1- (-1)^n}{2}= 0/2= 0[/itex] and[itex]\frac{1+ (-1)^n}{2}= 2/2= 1[/itex]. Then [itex]a_n= 0+ (n+1)^2= (n+1)^2[/itex].

When n is odd [itex](-1)^n= -1[/itex] so [itex]\frac{1- (-1)^n}{2}= 2/2= 1[/itex] and [itex]\frac{1- (-1)^n}{2}= 0[/itex]. Then [itex]a_n= 1+ 0(n+1)^2= 1[/itex].

Thanks!
 


arildno said:
There are many ways to do this, the most compact one is probably:

[tex]a_{n}=n^{(1+(-1)^{n})},n=1,2...[/tex]

Thank you very much!
 


Alexx1 said:
Can someone help me with this sequence?

1 4 1 16 1 36

If I only look at the odd numbers it's: n^2

But I don't know how they get to '1'..

The even members are [itex](2 n)^{2}, n = 1, 2, \ldots[/itex]. The odd ones are always 1.

So, we can write:

[tex] a_{n} = \left\{\begin{array}{rc}<br /> n^{2}, & n \, \mathrm{even} \\<br /> <br /> 1, & n \, \mathrm{odd}<br /> \end{array}\right.[/tex]

You can write it in this form:

[tex] a_{n} = \frac{n^{2} + 1}{2} + (-1)^{n} \, \frac{n^{2} - 1}{2} = \frac{1 - (-1)^{n}}{2} + n^{2} \, \frac{1 + (-1)^{n}}{2}[/tex]
 


Dickfore said:
The even members are [itex](2 n)^{2}, n = 1, 2, \ldots[/itex]. The odd ones are always 1.

So, we can write:

[tex] a_{n} = \left\{\begin{array}{rc}<br /> n^{2}, & n \, \mathrm{even} \\<br /> <br /> 1, & n \, \mathrm{odd}<br /> \end{array}\right.[/tex]

You can write it in this form:

[tex] a_{n} = \frac{n^{2} + 1}{2} + (-1)^{n} \, \frac{n^{2} - 1}{2} = \frac{1 - (-1)^{n}}{2} + n^{2} \, \frac{1 + (-1)^{n}}{2}[/tex]
Mine's better! :approve:
 


I'm not exactly sure, but I can think of a way to get back to the number by.. say...

[tex]\frac{1+(-1)^{n-1}}{2}[/tex]

In this formula, when n is odd, then we have 1, but when n is even we get 0.

So... we could use something like this

[tex](S(n/2)-1).\frac{(1+(-1)^{n-1})}{2}+1[/tex]

seems like it would be an appropriate formula for your problem, where S(n/2) is the sequence 4,16,36 for n=1,2,3.

So if you find a formula for the sequence S(n)=4,16,36 which I couldn't figure out (I'm pretty bad at sequences) then you can find a formula for the sequence you've shown by filling in S(n/2).