Forum archive 2000-2006

Michael Gage - Distributions.pm

Michael Gage - Distributions.pm

by Arnold Pizer -
Number of replies: 0
inactiveTopicDistributions.pm topic started 9/14/2004; 1:52:58 PM
last post 9/14/2004; 1:52:58 PM
userMichael Gage - Distributions.pm  blueArrow
9/14/2004; 1:52:58 PM (reads: 1699, responses: 0)


NAME

Statistics::Distributions - Perl module for calculating probabilities and critical values of common statistical distributions


SYNOPSIS

    use Statistics::Distributions;
    $uprob=uprob(-0.85);
print "upper probability of the u distribution: Q(u) = 1-G(u) (u=1.43) = $uprob";
    $tprob=tprob(3, 6.251);
print "upper probability of the t distribution: Q = 1-G (3 degrees of freedom
, t = 6.251) = $tprob";
    $chisprob=chisqrprob(3, 6.25);
print "upper probability of the chi-square distribution: Q = 1-G (3 degrees of
freedom, chi-squared = 6.25) = $chisprob";
    $fprob=fprob(3,5, .625);
print "upper probability of the F distribution: Q = 1-G (3 degrees of freedom
in numerator, 5 degrees of freedom in denominator, F $
    $u=udistr(.05);
print "u-crit (95th percentile = 0.05 level) = $u";
    $t=tdistr(1, .005);
print "t-crit (1 degree of freedom, 99.5th percentile = 0.005 level) =$t";
    $chis=chisqrdistr(2, .05);
print "Chi-squared-crit (2 degrees of freedom, 95th percentile = 0.05 level)
= $chis";
    $f=fdistr(1,3, .01);
print "F-crit (1 degree of freedom in numerator, 3 degrees of freedom in denominator,
99th percentile = 0.01 level) = $f";


DESCRIPTION

This Perl module calulates percentage points (6 significant digits) of the u (standard normal) distribution, the student's t distribution, the chi-square distribution and the F distribution.

It can also calculate the upper probability (6 significant digits) of the u (standard normal), the chi-square, the t and the F distribution.

These critical values are needed to perform statistical tests, like the u test, the t test, the chi-squared test, and the F test, and to calculate confidence intervals.

If you are interested in more precise algorithms you could look at: StatLib: http://lib.stat.cmu.edu/apstat/ Applied Statistics Algorithms by Griffiths, P. and Hill, I.D., Ellis Horwood: Chichester (1985)


AUTHOR

Michael Kospach, mike.perl@gmx.at Nice formating, simplification and bug repair by Matthias Trautner Kromann, mtk@id.cbs.dk

File path = /ww/webwork/pg/lib/Distributions.pm

<| Post or View Comments |>