Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Finite Periodicity of Pi

  1. Dec 7, 2006 #1
    If I take the first p digits of pi, is it possible that what I will have is the first p/2 digits repeated twice?

    For example, suppose pi = 3.14153141529485729487... and I took the first 10 digits and got 3141531415 which is 31415 repeated twice. Can this happen?
  2. jcsd
  3. Dec 7, 2006 #2
    It's conceivable that there exists irrational numbers where the first n digits in the decimal expansion repeat themselves and then after 2n digits, it goes non-repeating, but that's not true with pi, at least to a trillion digits.


    Infact, it's not just conceivable, it's pretty obvious!

    [tex]\sqrt{2} = 1.4 \ldots[/tex] Therefore [tex]0.123123 + \frac{\sqrt{2}}{10^{7}}[/tex] is irrational, but has the first three digits repeat.
    Last edited: Dec 7, 2006
  4. Dec 7, 2006 #3
    Yeah, I knew it wasn't true for arbitrary irrationals because the first example I thought of was (101010+pi) which is irrational and repeats 10 to start. I guess I was looking for existence of evidence that it happens in pi or if there's some way to prove that this never happens for some number.
  5. Dec 11, 2006 #4

    Gib Z

    User Avatar
    Homework Helper

    Pi can been calculated to i think its 14 trillion digits? No pattern found, but there are yet no proofs as to why, or even that each digit occurs infinity in the decimal expansion...
  6. Dec 12, 2006 #5


    User Avatar
    Science Advisor

    It is quite possible that for some n, the second n digits in [itex]\pi[/itex] are just the first n repeated but, if so, it must be some exceptionally large n. No one knows for sure but I would consider it unlikely.

    I notice you titled this "periodicity of [itex]\pi[/itex]" which implies a block of digits repeating over and over again, forever. Since [itex]\pi[/itex] is NOT rational, that certainly cannot happen.
  7. Dec 12, 2006 #6
    By "finite periodicity" I meant that the first n numbers are repeated a finite number of times, so it may appear periodic if you only look at the first k*n digits.
  8. Dec 12, 2006 #7


    User Avatar
    Science Advisor

    Ah. In that case, as everyone else has told you- possible, unknown, very unlikely! (Unlikely that the first n numbers repeat k times for some n, k. Highly likely that some set of n numbers repeats k times in a row but still unknown.)
  9. Jun 27, 2011 #8
    With the assumption of normality of pi, can we calculate the probability that this statement is true for some p?
    I think for a normal number the probability is less than 1/9.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook