13.07.2015 Views

Mathematical Methods for Physics and Engineering - Matematica.NET

Mathematical Methods for Physics and Engineering - Matematica.NET

Mathematical Methods for Physics and Engineering - Matematica.NET

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

PROBABILITYIf X <strong>and</strong> Y are both discrete RVs thenp(z) = ∑ i,jf(x i ,y j ), (30.60)where the sum extends over all values of i <strong>and</strong> j <strong>for</strong> which Z(x i ,y j )=z. Similarly,if X <strong>and</strong> Y are both continuous RVs then p(z) is found by requiring that∫∫p(z) dz = f(x, y) dx dy, (30.61)dSwhere dS is the infinitesimal area in the xy-plane lying between the curvesZ(x, y) =z <strong>and</strong> Z(x, y) =z + dz.◮Suppose X <strong>and</strong> Y are independent continuous r<strong>and</strong>om variables in the range −∞ to ∞,with PDFs g(x) <strong>and</strong> h(y) respectively. Obtain expressions <strong>for</strong> the PDFs of Z = X + Y <strong>and</strong>W = XY .Since X <strong>and</strong> Y are independent RVs, their joint PDF is simply f(x, y) =g(x)h(y). Thus,from (30.61), the PDF of the sum Z = X + Y is given by∫ ∞∫ z+dz−xp(z) dz = dx g(x) dy h(y)−∞z−x(∫ ∞)= g(x)h(z − x) dx dz.−∞Thus p(z) istheconvolution of the PDFs of g <strong>and</strong> h (i.e. p = g ∗ h, see subsection 13.1.7).In a similar way, the PDF of the product W = XY is given by∫ ∞∫ (w+dw)/|x|q(w) dw = dx g(x) dy h(y)−∞w/|x|(∫ ∞= g(x)h(w/x) dx )dw ◭−∞|x|The prescription (30.61) is readily generalised to functions of n r<strong>and</strong>om variablesZ = Z(X 1 ,X 2 ,...,X n ), in which case the infinitesimal ‘volume’ element dS is theregion in x 1 x 2 ···x n -space between the (hyper)surfaces Z(x 1 ,x 2 ,...,x n )=z <strong>and</strong>Z(x 1 ,x 2 ,...,x n )=z + dz. In practice, however, the integral is difficult to evaluate,since one is faced with the complicated geometrical problem of determining thelimits of integration. Fortunately, an alternative (<strong>and</strong> powerful) technique exists<strong>for</strong> evaluating integrals of this kind. One eliminates the geometrical problem byintegrating over all values of the variables x i without restriction, while shiftingthe constraint on the variables to the integr<strong>and</strong>. This is readily achieved bymultiplying the integr<strong>and</strong> by a function that equals unity in the infinitesimalregion dS <strong>and</strong> zero elsewhere. From the discussion of the Dirac delta function insubsection 13.1.3, we see that δ(Z(x 1 ,x 2 ,...,x n )−z) dz satisfies these requirements,<strong>and</strong> so in the most general case we have∫∫ ∫p(z) = ··· f(x 1 ,x 2 ,...,x n )δ(Z(x 1 ,x 2 ,...,x n ) − z) dx 1 dx 2 ...dx n ,(30.62)1154

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!