For a < b, the right-triangular distribution has density function
and the left-triangular distribution has density function
These distributions are denoted by RT(a, b) and LT(a, b), respectively.
(a) Show that if X ~ RT(0, 1), then Xʹ = a + (b – a)X ~ RT(a, b); verify the same relation between LT(0, 1) and LT(a, b). Thus it is sufficient to generate from RT(0, 1) and LT(0, 1).
(b) Show that if X ~ RT(0, 1), then 1 – X ~ LT(0, 1). Thus it is enough to restrict our attention further to generating from RT(0, 1).
(c) Derive the inverse-transform algorithm for generating from RT(0, l). Despite the result in (b), also derive the inverse-transform algorithm for generating directly from LT(0, l).
(d) As an alternative to the inverse-transform method, show that if U1 and U2 are IID U(0, 1) random variables, then max{U1, U2} ~ RT(0, 1). Do you think that this is better than the inverse-transform method? In what sense? (See Example 8.4.)
Example 8.4
We need at least 10 more requests to produce the solution.
0 / 10 have requested this problem solution
The more requests, the faster the answer.