Jacobian transformation, f_z(z) = d g / d (x, y) * (f_x(x), f

Jacobian transformation, f_z(z) = d g / d (x, y) * (f_x(x), f_y(y))

送交者: 潜伏九号于 2010-07-23, 20:35:47:

回答: 问个一般概率问题由 OO8 于 2010-07-23, 18:12:39:

= matrix of 2 marginal distribution and 2 join distribution.

partial g / partial (x, y) = J = Jacobian

e.g., if g is x+y, it is bivariate r.v.
if g is x/y, it may not exists.
if g is x*y ( x is normal and y is binary), it is mixture r.v.