Matrices whose entries are prime numbers

Discussion in 'Math Research' started by ss54, Feb 15, 2006.

  1. ss54

    ss54 Guest

    Let M be an n by n matrix whose entries are n^2 prime numbers. Is is
    true that this matrix is invertible?

    Simone Severini
     
    ss54, Feb 15, 2006
    #1
    1. Advertisements

  2. Take an n x n matrix whose all elements are 2. That certainly is not
    invertible! Even assuming distinct primes that's not true (although
    obviously holds for 2x2 matrices), take for example

    3 5 7
    11 13 17
    29 43 59

    It is composed of primes and singular.
     
    RobertSzefler, Feb 15, 2006
    #2
    1. Advertisements

  3. ss54

    G. A. Edgar Guest

    n=2, yes; n=3, no.

    A "prime triple" is the situation (p, p+2, p+6) where they
    are all primes. It is conjectured that there are infintely
    many prime triples. But in any case, there are at least three
    of them. Use them as rows in a 3x3 matrix to get a matrix
    with rank 2. For example:

    [ 5 7 11]
    [17 19 23]
    [41 43 47]

    has rank 2, and it thus not invertible.
     
    G. A. Edgar, Feb 15, 2006
    #3
  4. No. Consider a matrix all of whose entries are the same prime.
    Less trivially,

    ( 11 13 17 19 )
    ( 101 103 107 109 )
    ( 191 193 197 199 )
    ( 821 823 827 829 )

    has rank 2.
     
    Peter L. Montgomery, Feb 15, 2006
    #4
  5. ss54

    Rob Arthan Guest

    Here's a minimal counter-example that just uses the first 9 primes:

    3 2 5
    11 13 7
    19 17 23

    "Method": take the rows of the matrix to be triples of primes p, q, and r
    obeying a linear dependence relation Ap + Bq+ Cr for some A, B, C. Fiddle
    around with small A,B and C until you can get p, q and r really small, (A =
    3, B = -2 and C = -1 above).

    Regards,

    Rob.
     
    Rob Arthan, Feb 15, 2006
    #5
  6. ss54

    ss54 Guest

    Thanks for your kind replays.

    I think that when I wrote the post, I did not think enough. Of course,
    "distinct" was implicit.

    I need to modify the question:

    The elements of the matrix are consecutive primes.

    Specifically the examples I considered are the following:


    2 3
    5 7
    rank: 2

    2 3 5
    7 11 13
    17 19 23
    rank: 3

    2 3 5 7
    11 13 17 19
    23 29 31 37
    41 43 47 53
    rank: 4

    2 3 5 7 11
    13 17 19 23 29
    31 37 41 43 47
    53 59 61 67 71
    73 79 83 89 97
    rank: 5

    Probably I should go a bit further and try to find a counterexample...

    Thanks.
     
    ss54, Feb 16, 2006
    #6
  7. Gerald> A "prime triple" is the situation (p, p+2, p+6) where they are
    Gerald> all primes. It is conjectured that there are infintely many
    Gerald> prime triples.

    Green and Tao have announced a proof of the prime k-tuple conjecture
    -- see an article in the new Bulletin of the AMS. So this would
    extend to any n by n matrix.

    Victor
     
    Victor S. Miller, Feb 22, 2006
    #7
  8. Gerald> A "prime triple" is the situation (p, p+2, p+6) where they are
    Gerald> all primes. It is conjectured that there are infintely many
    Gerald> prime triples.

    Victor> Green and Tao have announced a proof of the prime k-tuple
    Victor> conjecture -- see an article in the new Bulletin of the AMS.
    Victor> So this would extend to any n by n matrix.

    I, of course, misspoke. While the Green and Tao result is a wonderful
    result (that there are arbitrarily long arithmetic progressions
    consisting only of primes), it is NOT the same as the prime k-tuple
    conjecture! Mea Culpa!

    Victor
     
    Victor S. Miller, Feb 22, 2006
    #8
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.