The gamma function is part of the formula defining the student's
t-distribution:

  y = Gamma * ((n+1/2)/Gamma * (n/2)*SQR(n*pi)) * (1+(t^2/n))exp -(n+1)/2

if that makes sense (sorry, I only have ascii on this machine).

n = df

Since df is a function of sample size (or sizes), the shape of the curve
changes as a function of df.

Mike


On Sun, 16 Feb 2003, Herman Rubin wrote:

> In article <[EMAIL PROTECTED]>,
> Dana Netherton  <[EMAIL PROTECTED]> wrote:
> >On Sat, 15 Feb 2003 01:27:34 GMT, [EMAIL PROTECTED] (john v
> >verkuilen) said ...
> >> Dana Netherton <[EMAIL PROTECTED]> writes:
>
> >> >So now my question is -- what's a "CDF Gamma" function
> >> >anyway?  How is it different from "a Gamma function"?
>
> The rest of what is quoted is below.  I believe it will
> be clear why.
>
> Many distributions are named for the functions used to obtain
> or describe their probabilities or densities.  For example,
> the geometric distribution has its probabilities forming a
> geometric series.  The binomial distribution has the terms
> of a binomial expansion, as does the multinomial.  The
> hypergeometric distribution has terms proportional to the
> coefficients in the expansion of a hypergeometric function.
>
> The exponential distribution has an exponential function
> for its density.  The double exponential distribution as
> exponentials going both ways from the center.  The Beta and
> Gamma distributions have as their densities the functions
> which when integrated give the Beta and Gamma functions.
>
> >> The Gamma function is an integral which generalizes the factorial since
> >> Gamma(x) = (x-1)! for whole numbers x > 1.  It is typically defined over
> >> non-integer values of x, at least for the purposes of probability theory.
> >> You might want to see http://www.math2.org/math/integrals/more/gammafun.htm.
>
> >> The CDF Gamma is the CDF of the gamma distribution, which is based on the
> >> gamma function.  It should be discussed in any decent probability theory
> >> book.  The Gamma distribution is a commonly used waiting-time distribution.
> >> I recall there's a connection between it and the Weibull but I can't remember
> >> the details and don't have a reference handy.
>
> >Hmm ... so the gamma distribution function is so-called
> >because it uses the gamma function inside it? (Yes, I
> >remember seeing the gamma function inside its defining
> >formula)
> --
> This address is for information only.  I do not claim that these views
> are those of the Statistics Department or of Purdue University.
> Herman Rubin, Deptartment of Statistics, Purdue University
> [EMAIL PROTECTED]         Phone: (765)494-6054   FAX: (765)494-0558
> .
> .
> =================================================================
> Instructions for joining and leaving this list, remarks about the
> problem of INAPPROPRIATE MESSAGES, and archives are available at:
> .                  http://jse.stat.ncsu.edu/                    .
> =================================================================
>
.
.
=================================================================
Instructions for joining and leaving this list, remarks about the
problem of INAPPROPRIATE MESSAGES, and archives are available at:
.                  http://jse.stat.ncsu.edu/                    .
=================================================================

Reply via email to