The R Project SVN R

Rev

Rev 88603 | Show entire file | Ignore whitespace | Details | Blame | Last modification | View Log | RSS feed

Rev 88603 Rev 90186
Line 58... Line 58...
58 
  \eqn{F(x) = 1 - \exp(-{(x/\sigma)}^a)}{F(x) = 1 - exp(- (x/b)^a)}
 58 
  \eqn{F(x) = 1 - \exp(-{(x/\sigma)}^a)}{F(x) = 1 - exp(- (x/b)^a)}
59 
  on \eqn{x > 0}, the
 59 
  on \eqn{x > 0}, the
60 
  mean is \eqn{E(X) = \sigma \Gamma(1 + 1/a)}{E(X) = b \Gamma(1 + 1/a)}, and
 60 
  mean is \eqn{E(X) = \sigma \Gamma(1 + 1/a)}{E(X) = b \Gamma(1 + 1/a)}, and
61 
  the variance is \eqn{Var(X) = \sigma^2(\Gamma(1 + 2/a)-(\Gamma(1 + 1/a))^2)}{
 61 
  the variance is \eqn{Var(X) = \sigma^2(\Gamma(1 + 2/a)-(\Gamma(1 + 1/a))^2)}{
62 
                       Var(X) = b^2 * (\Gamma(1 + 2/a) - (\Gamma(1 + 1/a))^2)}.
 62 
                       Var(X) = b^2 * (\Gamma(1 + 2/a) - (\Gamma(1 + 1/a))^2)}.
- 
 
 63 
 
- 
 
 64 
  The Weibull distribution is defined only for \eqn{a > 0}, but
- 
 
 65 
  these functions accept the argument \code{shape = 0}. In this case, the
- 
 
 66 
  return value is determined by the limiting behaviour as
- 
 
 67 
  \eqn{a \to 0}{a -> 0}.
63 
}
 68 
}
64 
\note{
 69 
\note{
65 
  The cumulative hazard \eqn{H(t) = - \log(1 - F(t))}{H(t) = - log(1 - F(t))}
 70 
  The cumulative hazard \eqn{H(t) = - \log(1 - F(t))}{H(t) = - log(1 - F(t))}
66 
  is
 71 
  is
67 
\preformatted{
 72 
\preformatted{