[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference rusure::math

Title:Mathematics at DEC
Moderator:RUSURE::EDP
Created:Mon Feb 03 1986
Last Modified:Fri Jun 06 1997
Last Successful Update:Fri Jun 06 1997
Number of topics:2083
Total number of notes:14613

1125.0. "proof" by PARITY::MOORADIAN () Tue Sep 19 1989 09:37

    problem:
    --------
    
    
    show that
    
    
    	--
       /     /
     --     d (ax-b)dx=1/lal
    
    
    (limits of integration are (-infinity to +infinity)
    
           /
    where d = delta and dx is derivitive
    
    lal= absolute value of a
    
    
T.RTitleUserPersonal
Name
DateLines
1125.1HPSTEK::XIAIn my beginning is my end.Tue Sep 19 1989 11:143
    What is delta?
    
    Eugene
1125.2A sketchof the proof floows:COOKIE::PBERGHPeter Bergh, DTN 523-3007Tue Sep 19 1989 12:0945
                    <<< Note 1125.0 by PARITY::MOORADIAN >>>
                                   -< proof >-

    >> problem:
    >> --------
    
    
    >> show that
    
    
    >> 	--
    >>    /     /
    >>  --     d (ax-b)dx=1/lal
    
    
    >> (limits of integration are (-infinity to +infinity)
    
    >>        /
    >> where d = delta and dx is derivitive
    
    >> lal= absolute value of a
    
    By "delta", I assume that you refer to Dirac's delta distribution,
    defined as that distribution (on the real line) that transforms a
    function to its value at the origin.
    
    A "proof" is done by treating delta as a nice (e.g., infinitely
    differentiable function) and simply making the variable transformation
    ax-b --> z; the result falls out that way (the absolute value of a
    follows from that you have to change the direction of the integration
    if a<0).
    
    This technique can be made stringent by constructing some sequence of
    infinitely differentiable functions that "converge to delta", i.e.,
    such that the (convolution) integral of the function f and members of
    the sequence converge to f(0).  By using, e.g., dominated convergence
    (Lebesgue's theorem on ...), you should be able to prove what you need
    to.
    
    A functional-analysis-based approach to a stringent proof can be found
    in any reasonable book on the theory of distributions (e.g., Laurent
    Schwartz's original book on the subject: "Theorie des distributions",
    published in Paris in the late fifties or early sixties).