Page 1 of 3 123 LastLast
Results 1 to 30 of 63

Thread: Does Pi contain Pi?

  1. #1
    Join Date
    Apr 2010
    Posts
    304

    Does Pi contain Pi?

    Does the decimal expansion of Pi contain the decimal expansion of Pi?
    and so on...
    and so on....

  2. #2
    Join Date
    Oct 2009
    Location
    a long way away
    Posts
    9,922
    Well, only once.

  3. #3
    Join Date
    Feb 2014
    Posts
    182
    Quote Originally Posted by Strange View Post
    Well, only once.
    Took me a second.

  4. #4
    Join Date
    Jul 2005
    Posts
    14,622
    Quote Originally Posted by Strange View Post
    Well, only once.
    And no more than that.
    If pi contained a subset of digits identical to its own decimal expansion, then the sequence of digits before the subset would be repeated within the subset; and the subset would contain a similar sequence of digits repeated within a further subset ... And so on. Which would mean that pi was a repeating decimal, which would mean it could be expressed as a ratio, which would mean it was a rational number. Which it isn't. Therefore pi cannot contain a subset of digits that repeat its own decimal expansion.

    Grant Hutchison

  5. #5
    Join Date
    Jan 2002
    Location
    The Valley of the Sun
    Posts
    8,773
    Maybe beyond some point in the digits of pi, every other digit is a digit of pi, in order. Would that count as pi containing itself?

  6. #6
    Join Date
    Mar 2004
    Posts
    15,692
    Good point. If you're willing to just encounter the digits, skipping over the digits that don't match rather than regularly skipping a fixed-size gap, then you will be able to re-find the digits of pi, just not consecutively.

    Containment can have different levels. My own fetish number (down there in my signature, the Thue-Morse sequence), an infinite non-repeating sequence, does contain itself in a way. Regard the pattern of "01" and "10" pairs you see, or the "0110" and "1001" quads. Both (and infinitely more) appear in the same pattern. If you change every 01 to zero and every 10 to one, you get the same infinite sequence.
    0 1 1 0 1 0 0 1 1 0 0 1 0 1 1 0 1 0 0 1 0 1 1 0 0 1 1 0 1 0 0 1 1 0 0 1 0 1 1 0 0 1 1 0 1 0 0 1 0 1 1 0 1 0 0 1 1 0 0 1 0 1 1 0 ...
    [COLOR="#f9f9f9"]Testing... 1... 2... 3...[/COLOR]

  7. #7
    Join Date
    Mar 2004
    Posts
    15,692
    Oh, I neglected to mention the more obvious containment of the Thue-Morse (see signature) sequence within itself. It is exactly as Chuck wondered about pi. From Thue-Morse, take every other digit. You get Thue-Morse. Take every 4th digit. Take every 8th, or 16th. You get Thue-Morse again. Every 32nd, etc.

    Again, this is a nonrepeating infinite sequence. And it's so nonrepeating that it completely avoids any triples, any repetition of one group three times: no 000, no 010101, no 101101101, etc. Way cooler than a number deserves to be.
    0 1 1 0 1 0 0 1 1 0 0 1 0 1 1 0 1 0 0 1 0 1 1 0 0 1 1 0 1 0 0 1 1 0 0 1 0 1 1 0 0 1 1 0 1 0 0 1 0 1 1 0 1 0 0 1 1 0 0 1 0 1 1 0 ...
    [COLOR="#f9f9f9"]Testing... 1... 2... 3...[/COLOR]

  8. #8
    Join Date
    May 2007
    Location
    Earth
    Posts
    9,734
    Quote Originally Posted by grant hutchison View Post
    And no more than that.
    If pi contained a subset of digits identical to its own decimal expansion, then the sequence of digits before the subset would be repeated within the subset; and the subset would contain a similar sequence of digits repeated within a further subset ... And so on. Which would mean that pi was a repeating decimal, which would mean it could be expressed as a ratio, which would mean it was a rational number. Which it isn't. Therefore pi cannot contain a subset of digits that repeat its own decimal expansion.

    Grant Hutchison
    Infinities do strange things. The infinitely long expansion of pi could contain an infinitely long expansion of pi...

    I'll go take my meds now.

    Information about American English usage here and here. Floating point issues? Please read this before posting.

    How do things fly? This explains it all.

    Actually they can't: "Heavier-than-air flying machines are impossible." - Lord Kelvin, president, Royal Society, 1895.



  9. #9
    Join Date
    Jul 2005
    Posts
    14,622
    Quote Originally Posted by swampyankee View Post
    Infinities do strange things. The infinitely long expansion of pi could contain an infinitely long expansion of pi...
    Well, that's what I was talking about - an infinite subset, which in turn must contain its own identical infinite subset, which in turn contains another identical infinite subset, and so on. That would make it a repeating decimal. But pi is an irrational number, which means its non-repeating. Therefore it can't contain such nested identical infinite subsets.

    Grant Hutchison

  10. #10
    Join Date
    Jul 2005
    Posts
    14,622
    Putting what I've been saying on a mathematical footing:

    If pi contains itself, then we could multiply it by 10n (for some integer value of n, corresponding to the number of decimal places at which the contained copy starts) and recover pi, like this:

    pi * 10n = x + pi (where x is an integer that contains all the digits before the contained copy)

    Rearranging gives us:

    pi = x/(10n-1)

    Which is a ratio of two integers, which conflicts with our knowledge that pi is irrational.

    Grant Hutchison

  11. #11
    Join Date
    May 2003
    Posts
    5,770
    And of course if we take Strange's suggestion, that it shows up once, that would be the case where x and n are both zero, which are the only values for which grant's equation works.
    Conserve energy. Commute with the Hamiltonian.

  12. #12
    Join Date
    Jul 2005
    Posts
    14,622
    Quote Originally Posted by Grey View Post
    And of course if we take Strange's suggestion, that it shows up once, that would be the case where x and n are both zero, which are the only values for which grant's equation works.


    Grant Hutchison

  13. #13
    Join Date
    May 2005
    Posts
    7,675
    Quote Originally Posted by Grey View Post
    And of course if we take Strange's suggestion, that it shows up once, that would be the case where x and n are both zero, which are the only values for which grant's equation works.
    His first equation, not his second

  14. #14
    Join Date
    Jul 2005
    Posts
    14,622
    Quote Originally Posted by grapes View Post
    His first equation, not his second
    I was applauding the creative use of the second equation.

    Grant Hutchison

  15. #15
    Join Date
    Mar 2009
    Posts
    1,561
    Quote Originally Posted by swampyankee View Post
    Infinities do strange things. The infinitely long expansion of pi could contain an infinitely long expansion of pi...

    In this universe Pi is an irrational and a transcendental number.

    If Pi had been exactly equal to 3, would the history of the development of mathematics & man's perception of infinity have plaued out significantly differently in any way from OTL?
    Last edited by wd40; 2017-Oct-01 at 03:36 PM.

  16. #16
    Join Date
    May 2007
    Location
    Earth
    Posts
    9,734
    Quote Originally Posted by wd40 View Post
    In this universe Pi is an irrational and a transcendental number.

    If Pi had been exactly equal to 3, would the history of the development of mathematics out significantly different in any way from OTL?
    I'm quite aware of that; it can't be represented by a finite expansion. Infinities behave strangely: if I number each one of the real numbers between 0 and 1 with a whole number, there would be a one-to-one correlation between 0<x<1 and all whole numbers. This can be repeated between any two arbitrary real numbers that aren't equal, e.g., 0.5<x<0.6.

    Information about American English usage here and here. Floating point issues? Please read this before posting.

    How do things fly? This explains it all.

    Actually they can't: "Heavier-than-air flying machines are impossible." - Lord Kelvin, president, Royal Society, 1895.



  17. #17
    Join Date
    Jul 2005
    Posts
    14,622
    Quote Originally Posted by swampyankee View Post
    I'm quite aware of that; it can't be represented by a finite expansion. Infinities behave strangely: if I number each one of the real numbers between 0 and 1 with a whole number ...
    Stranger than that, even.
    The set of real numbers between 0 and 1 is uncountably infinite, so you can't put it into a one-to-one relationship with the set of whole numbers, which is countably infinite.

    Grant Hutchison

  18. #18
    Join Date
    Mar 2004
    Posts
    15,692
    Quote Originally Posted by wd40 View Post
    If Pi had been exactly equal to 3 [...]
    ... then we'd be using a different symbol to represent the ratio of a circle's circumference and diameter.
    0 1 1 0 1 0 0 1 1 0 0 1 0 1 1 0 1 0 0 1 0 1 1 0 0 1 1 0 1 0 0 1 1 0 0 1 0 1 1 0 0 1 1 0 1 0 0 1 0 1 1 0 1 0 0 1 1 0 0 1 0 1 1 0 ...
    [COLOR="#f9f9f9"]Testing... 1... 2... 3...[/COLOR]

  19. #19
    Join Date
    Jun 2005
    Posts
    1,742
    Not repeating but terminating and in part repeating in bases of pi and its roots:

    https://www.wolframalpha.com/input/?...e+(sqrt(pi,6))

    There probably is a way to make the digits repeat, but the base would probably be unrepresentable using existing mathematical notations.

    Is there a base n where pi = 1010?
    Probably, but the n would be an irrational number unrepresentable in full but calculable to any number of given decimal points.

    Corrections are welcome.

    ETA Perhaps:
    http://m.wolframalpha.com/input/?i=s...80%3Dn%5E3%2Bn
    Last edited by a1call; 2017-Oct-02 at 04:57 AM.

  20. #20
    Join Date
    Dec 2005
    Posts
    1,353
    I've always loved the fact that

    e^(pi*i)=-1

    I love irrational transcendental numbers too. Somewhere in pi there are a billion zeroes in a row.

    And if you assign each triplet of digits as the ASCII codes for characters, pi somewhere contains this post, your post, in fact the entire forum, plus the DNA of everything living creature on the Earth, plus our entire history and every book ever written and every conversation ever engaged in. An infinite number of times.

    clop

  21. #21
    Join Date
    Jan 2002
    Location
    The Valley of the Sun
    Posts
    8,773
    It should be impossible to copyright anything since the material previously existed in pi and is therefor not an original work.

  22. #22
    Join Date
    Jun 2005
    Posts
    12,393
    Quote Originally Posted by clop View Post
    And if you assign each triplet of digits as the ASCII codes for characters, pi somewhere contains this post, your post, in fact the entire forum, plus the DNA of everything living creature on the Earth, plus our entire history and every book ever written and every conversation ever engaged in. An infinite number of times.
    I'm going to sue you for making my brain hurt.
    As above, so below

  23. #23
    Join Date
    Jul 2005
    Posts
    14,622
    Quote Originally Posted by clop View Post
    And if you assign each triplet of digits as the ASCII codes for characters, pi somewhere contains this post, your post, in fact the entire forum, plus the DNA of everything living creature on the Earth, plus our entire history and every book ever written and every conversation ever engaged in. An infinite number of times.
    We don't actually know if that's true. Just because it's infinite and non-repeating doesn't mean it has to contain every possible sequence of digits.

    Grant Hutchison

  24. #24
    Join Date
    Mar 2004
    Posts
    15,692
    Quote Originally Posted by Chuck View Post
    It should be impossible to copyright anything since the material previously existed in pi and is therefor not an original work.
    Fortunately, that does not disqualify a copyright. If two people create an original expression of the same idea in the same way, idependently, they both may own it.

    What could be disqualified is a non-creative copying of some part of pi. But that would have to be shown, not assumed.
    0 1 1 0 1 0 0 1 1 0 0 1 0 1 1 0 1 0 0 1 0 1 1 0 0 1 1 0 1 0 0 1 1 0 0 1 0 1 1 0 0 1 1 0 1 0 0 1 0 1 1 0 1 0 0 1 1 0 0 1 0 1 1 0 ...
    [COLOR="#f9f9f9"]Testing... 1... 2... 3...[/COLOR]

  25. #25
    Join Date
    Mar 2008
    Posts
    5,327
    Quote Originally Posted by grant hutchison View Post
    We don't actually know if that's true. Just because it's infinite and non-repeating doesn't mean it has to contain every possible sequence of digits.

    Grant Hutchison
    This kind of stuff makes my head hurt...

    We can't know for sure either way, I get that, I think... But would it not be sensible to assume that it must? Since that its non repeating and is infinite then should we assume that every possible sequence of digits is in there? Or is it because the possible sequence of digits is infinite also which gives it the ambiguity?

  26. #26
    Join Date
    Jul 2005
    Posts
    14,622
    Quote Originally Posted by cosmocrazy View Post
    This kind of stuff makes my head hurt...

    We can't know for sure either way, I get that, I think... But would it not be sensible to assume that it must? Since that its non repeating and is infinite then should we assume that every possible sequence of digits is in there? Or is it because the possible sequence of digits is infinite also which gives it the ambiguity?
    I think mathematicians believe it's likely that it does, but the last time I looked no-one had proved that it does. We don't even know if all digits 0 to 9 occur an infinite number of times in the decimal expansion of pi. It might just stop having nines at some point, for instance, leaving us with only a finite number of nines.
    And it's trivially easy to come up with an infinite, non-repeating sequence that doesn't contain all finite sequences - 0110001111000001111110000000... for instance.

    Grant Hutchison

  27. #27
    Join Date
    May 2005
    Posts
    7,675
    Quote Originally Posted by cosmocrazy View Post
    This kind of stuff makes my head hurt...

    We can't know for sure either way, I get that, I think... But would it not be sensible to assume that it must? Since that its non repeating and is infinite then should we assume that every possible sequence of digits is in there?
    It's easy to find counter examples. For instance, Liouville's constant is transcendental, irrational of course, infinite and non-repeating, but has only zeroes, and isolated single digits of 1, except for the first two digits.
    Or is it because the possible sequence of digits is infinite also which gives it the ambiguity?
    ETA:
    Quote Originally Posted by grant hutchison View Post
    I think mathematicians believe it's likely that it does, but the last time I looked no-one had proved that it does. We don't even know if all digits 0 to 9 occur an infinite number of times in the decimal expansion of pi. It might just stop having nines at some point, for instance, leaving us with only a finite number of nines.
    And it's trivially easy to come up with an infinite, non-repeating sequence that doesn't contain all finite sequences - 0110001111000001111110000000... for instance.
    Similarly, Liouville's constant is .1100010000000000100000... where the nth 1 is at the n! position.

  28. #28
    Join Date
    May 2003
    Posts
    5,770
    In mathematics, this property is called normality. A normal number is a real number in which every sequence of digits occurs with the same frequency as every other possible sequence of digits with the same length. It's possible to prove that "almost all" (that term has a specific mathematical meaning in this case) real numbers are normal, but it turns out to be tantalizingly hard to show that any specific real number is or is not normal. As grant points out, although most mathematicians suspect that pi is normal, it has not been proven to be so.
    Conserve energy. Commute with the Hamiltonian.

  29. #29
    Join Date
    Jul 2005
    Posts
    14,622
    Quote Originally Posted by Grey View Post
    In mathematics, this property is called normality.
    Ah, thank you! I've been trying to remember that word all afternoon. I got stuck with the words complete and exhaustive in my head, and couldn't seem to think around them.

    Grant Hutchison

  30. #30
    Join Date
    Jun 2005
    Posts
    1,742
    Quote Originally Posted by Grey View Post
    In mathematics, this property is called normality. A normal number is a real number in which every sequence of digits occurs with the same frequency as every other possible sequence of digits with the same length. It's possible to prove that "almost all" (that term has a specific mathematical meaning in this case) real numbers are normal, but it turns out to be tantalizingly hard to show that any specific real number is or is not normal. As grant points out, although most mathematicians suspect that pi is normal, it has not been proven to be so.
    The math involved in normal-numbers is way beyond my IQ, but with my naive interpretation, it should be possible to exist a not-normal number that contains any given finite sequence of digits, infinite times, just not with a uniform density. Same way a loaded Die will land 100 consecutive 1s, just much more rarely than it lands with 100 consecutive 6s.
    There might be sequences which never occur in pi, but not being normal may or may not be relevant.
    Corrections/insights are appreciated.
    Last edited by a1call; 2017-Oct-04 at 12:56 AM.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •