Derivation of P.D.F. from distribution function

1 reply · 2K views · last reply
WMDhamnekar
MHB
Messages
382
Reaction score
31
TL;DR
If ## F_{X_k}(x) = p(X \leq x) = \displaystyle\sum_{j=k}^n \binom{n}{j} F^j(x)(1-F(x))^{n-j}, -\infty
< x < \infty ## then how to prove ##f_{X_k} (x) =\frac{n!}{(k-1)!(n-k)!}f(x) F^{k-1}(x)(1-F(x))^{n-k}##
Author computed ##f_{X_k}(x)## as follows but I don't understand it. Would any member explain me the following computations?
Handwritten-style printed derivation of the density f_{X_k}(x) as a double sum over j of n!/((j-1)!(n-j)!) terms, then cancellation to n!/((k-1)!(n-k)!) f(x)F^{k-1}(x)(1-F(x))^{n-k} for -infinity < x < infinity.
 
WMDhamnekar said:
Summary: If ## F_{X_k}(x) = p(X \leq x) = \displaystyle\sum_{j=k}^n \binom{n}{j} F^j(x)(1-F(x))^{n-j}, -\infty
< x < \infty ## then how to prove ##f_{X_k} (x) =\frac{n!}{(k-1)!(n-k)!}f(x) F^{k-1}(x)(1-F(x))^{n-k}##

Author computed ##f_{X_k}(x)## as follows but I don't understand it. Would any member explain me the following computations?
Printed mathematical derivation of the density of the k-th order statistic, showing sums with n!/((j-1)!(n-j)!) terms that cancel to f(x)F^(k-1)(x)(1-F(x))^(n-k) times n!/((k-1)!(n-k)!), valid for -inf < x < inf.
I tag this question as "SOLVED". I understood all the computations. Thanks.
 
Last edited: