Change of Variables
In one dimension
Let be a continuous r.v. with PDF , and let , where is differentiable and strictly increasing (or strictly decreasing). Then the PDF of is given by
where . The support of is all with in the support of .
Tip
Do not forget specifying the support of
General case
Let be a continuous random vector with joint PDF . Let be an invertible function, where and are open subsets of , contains the support of , and is the range of .
Let . and mirror this by letting . Since is invertible, we also have and .
Suppose that all the partial derivatives exist and are continuous, so we can form the Jacobian matrix
Also assume that the determinant of this Jacobian matrix is never . Then the joint PDF of is
This relationship means that
Convolution
Let and are independent, then the PDF of is the convolution of :
Regenerative Property of Binomial Distribution
For Binomial Distribution, let , , we compute the PDF of
Therefore, . This is a discrete convolution
Regenerative Property of Poisson Distribution
For Poisson Distribution, let , then if are independent, we have
Tip
By replacing by , the convolution formula can also be used for
Product
Let the PDF of be , then the PDF of is
This is by
Therefore
Division
Let the PDF of be , then the PDF of is
This is by
Therefore