Chapter_50

# vn1 j where s1 v1 s1 v2 s2 v2 s1 vn1 j

This preview shows page 1. Sign up to view the full content.

This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: t; sn+1 < ∞ s1 < s2 < . . . < sn+1 = sn+1 . ⇒ s1 = v1 vn+1 s2 . . . = v2 vn+1 sn = vn vn+1 sn+1 = vn+1 fV (v1 , v2 , . . . , vn+1 ) = fS (v1 vn+1 , v2 vn+1 , . . . , vn+1 )|J | where ∂ s1 ∂v1 ··· ∂s1 ∂v2 ∂s2 ∂v2 ··· ∂s1 ∂vn+1 |J | = ∂s2 ∂v1 ∂s2 ∂vn+1 ··· . . . ∂sn+1 ∂v1 ∂sn+1 ∂v2 . . . ∂sn+1 ∂vn+1 vn+1 = 0 0 ··· 0 0 . . . vn+1 0 ··· 0 . . . v1 v2 v3 · · · 1 63 n = vn+1 n ⇒ fV (v1 , v2 , . . . , vn+1 ) = exp (−vn+1 ) vn+1 0 < vn+1 < ∞ v1 < v2 < . . . < vn+1 ∞ f V ( v1 , v 2 , . . . , v n ) = 0 n exp (−vn+1 ) vn+1 dvn+1 Successive integration by parts gives: fV ( v1 , v 2 , . . . , v n ) = n ! for v1 < v2 < . . . vn . We show that this is precisely the joint pdf of n ordered uniforms. Given X1 , X2 , . . . , Xn ∼ fX (x) then the joint pdf of the order statistics is fX(1) ,...,X(n) (x(1) , x(2) , . . . , x(n) ) = n!fX (x(1) )fX (x(2) ) . . . fX (x(n) ) Intuition: there are n! ways of gettin...
View Full Document

## This document was uploaded on 03/25/2014.

Ask a homework question - tutors are online