Results 1 to 7 of 7

Math Help - Evaluate the sum

  1. #1
    Member
    Joined
    Nov 2006
    Posts
    152

    Evaluate the sum

    Evaluate the sum


     [log_21]+[log_22]+[log_23]+...+[log_22002].
    Follow Math Help Forum on Facebook and Google+

  2. #2
    Senior Member Peritus's Avatar
    Joined
    Nov 2007
    Posts
    397
    using the well known rule:

    loga + logb = log(ab)

    we get:

    [log_21]+[log_22]+[log_23]+...+[log_22002] = log_21*2*3*...*2001*2002 = log_22002!
    Follow Math Help Forum on Facebook and Google+

  3. #3
    is up to his old tricks again! Jhevon's Avatar
    Joined
    Feb 2007
    From
    New York, USA
    Posts
    11,663
    Thanks
    3
    Quote Originally Posted by Peritus View Post
    using the well known rule:

    loga + logb = log(ab)

    we get:

    [log_21]+[log_22]+[log_23]+...+[log_22002] = log_21*2*3*...*2001*2002 = log_22002!
    i thought so too. but i figured that was too easy. isn't there a way to get an answer without logs? of course it would be a huge number (my calculator can't do log 2002! and i doubt anyone's calculator can), but perhaps we can get it in standard form
    Follow Math Help Forum on Facebook and Google+

  4. #4
    Forum Admin topsquark's Avatar
    Joined
    Jan 2006
    From
    Wellsville, NY
    Posts
    9,939
    Thanks
    338
    Awards
    1
    Quote Originally Posted by Jhevon View Post
    i thought so too. but i figured that was too easy. isn't there a way to get an answer without logs? of course it would be a huge number (my calculator can't do log 2002! and i doubt anyone's calculator can), but perhaps we can get it in standard form
    Stirling's approximation says that, for large n:

    n! \approx \sqrt{2 \pi n} \frac{n^n}{e^n}

    So
    ln(n!) \approx \frac{1}{2}ln(n) + n~ln(n) - n
    (ignoring the relatively small constant term.)

    So using the change of base formula:
    log_2(2002!) \approx \frac{ \frac{1}{2}ln(2002) + 2002 \cdot ln(2002) - 2002}{ln(2)} \approx  19073.6

    -Dan
    Follow Math Help Forum on Facebook and Google+

  5. #5
    is up to his old tricks again! Jhevon's Avatar
    Joined
    Feb 2007
    From
    New York, USA
    Posts
    11,663
    Thanks
    3
    Quote Originally Posted by topsquark View Post
    Stirling's approximation says that, for large n:

    n! \approx \sqrt{2 \pi n} \frac{n^n}{e^n}

    So
    ln(n!) \approx \frac{1}{2}ln(n) + n~ln(n) - n
    (ignoring the relatively small constant term.)

    So using the change of base formula:
    log_2(2002!) \approx \frac{ \frac{1}{2}ln(2002) + 2002 \cdot ln(2002) - 2002}{ln(2)} \approx 19073.6

    -Dan
    thanks.

    you know what the sad thing is? i learned Stirling's approximation in class just the other day, i should have thought of that. (it was in a probability class, i hate probability by the way)


    that number seems a LOT smaller than i thought it would be. why does my calculator have problems spitting that answer out, or anything close to it
    Follow Math Help Forum on Facebook and Google+

  6. #6
    Forum Admin topsquark's Avatar
    Joined
    Jan 2006
    From
    Wellsville, NY
    Posts
    9,939
    Thanks
    338
    Awards
    1
    Quote Originally Posted by Jhevon View Post
    thanks.

    you know what the sad thing is? i learned Stirling's approximation in class just the other day, i should have thought of that. (it was in a probability class, i hate probability by the way)


    that number seems a LOT smaller than i thought it would be. why does my calculator have problems spitting that answer out, or anything close to it
    If you ever take a Statistical Mechanics course you will forever think of the Stirling approximation. Trust me!

    Recall that you are looking at the logarithm of a number. So if log_x \approx 20000 then we are talking of an x around 2^{20000}, a not insignificant number. This is why they invented logs in the first place: early Mathematicians had the TI - 1, which vaguely looks like an abacus with a plug. They couldn't handle multiplication of big numbers. Logarithms reduce the size of the number you are working with, so you can tabulate them easier. (Then they had the HP - 1, thus losing the ability to even count, and we entered the Dark Ages.)

    -Dan
    Follow Math Help Forum on Facebook and Google+

  7. #7
    Super Member PaulRS's Avatar
    Joined
    Oct 2007
    Posts
    571
    I was wondering whether [x] denotes floor function or something like that, because it's a bit strange that perash used it for each term.

    It is more interesting if it is the floor function

    If it really denotes floor function, to calculate the sum, we should consider
    \sum_{k=1}^{n}{k\cdot{2^k}}

    \sum_{k=1}^{n}{k\cdot{x^k}}=x\cdot{\left(\sum_{k=0  }^{n}{x^k}\right)'}

    \sum_{k=1}^{n}{k\cdot{x^k}}=x\cdot{\left(\frac{x^{  n+1}-1}{x-1}\right)'}=x\cdot{\left(\frac{(n+1)x^{n}(x-1)-x^{n+1}+1}{(x-1)^2}\right)}

    \sum_{k=1}^{n}{k\cdot{2^k}}=(n+1)\cdot{2^{n+1}}-2^{n+2}+2

    Because our sum is: S=\sum_{k=1}^{10}{k\cdot{2^k}}-(2047-2002)\cdot{10}

    In general
    \sum_{k=1}^{n}{[\log_2(k)]}=\sum_{k=1}^{[\log_2(n)]}{k\cdot{2^k}}-(2^{[\log_2(n)]+1}-1-n)\cdot{[\log_2(n)]}

    Then: \sum_{k=1}^{n}{[\log_2(k)]}=([\log_2(n)]+1)\cdot{2^{[\log_2(n)]+1}}-2^{[\log_2(n)]+2}+2-(2^{[\log_2(n)]+1}-1-n)\cdot{[\log_2(n)]}
    Which is:
    \sum_{k=1}^{n}{[\log_2(k)]}=-2^{[\log_2(n)]+1}+2+(n+1)\cdot{[\log_2(n)]}
    Follow Math Help Forum on Facebook and Google+

Similar Math Help Forum Discussions

  1. evaluate
    Posted in the Algebra Forum
    Replies: 4
    Last Post: April 18th 2010, 12:16 PM
  2. evaluate
    Posted in the Calculus Forum
    Replies: 3
    Last Post: November 21st 2009, 03:26 AM
  3. Evaluate
    Posted in the Pre-Calculus Forum
    Replies: 1
    Last Post: January 26th 2009, 04:56 AM
  4. how do i evaluate...
    Posted in the Trigonometry Forum
    Replies: 3
    Last Post: September 4th 2008, 10:55 PM
  5. Please evaluate
    Posted in the Pre-Calculus Forum
    Replies: 3
    Last Post: December 14th 2007, 04:36 PM

Search Tags


/mathhelpforum @mathhelpforum