Statistics::OLS
- perform ordinary least squares and other bivariate statistics, v 0.07.
I wrote Statistics::OLS to perform Ordinary Least Squares (linear curve
fitting) on two dimensional data: y = a + bx. The other simple
statistical module I found on CPAN (Statistics::Descriptive) is
designed for univariate analysis. It accomodates OLS, but somewhat
inflexibly and without rich bivariate statistics. Nevertheless, it
might make sense to fold OLS into that module or a supermodule someday.
Statistics::OLS computes the estimated slope and intercept of the
regression line, their T-statistics, R squared, standard error of the
regression and the Durbin-Watson statistic. It can also return the
residuals.
It is pretty simple to do two dimensional least squares, but much
harder to do multiple regression, so Statistics::OLS is unlikely ever
to work with multiple independent variables. The
Statistics::Descriptive module performs useful univariate
statistics and is well tested. The Perl Data Language (see CPAN) may
also prove useful for statistics. For a heavy duty, free, and open
source statistical analysis language and environment, see the R
project at http://www.R-project.org
This is beta code, and has not been extensively tested. It has worked
on a few published datasets. Feedback is welcome, particularly if you
notice an error or try it with known results that are not reproduced
correctly.
Installation:
Untar and unzip the distribution and cd into its directory. Then the
usual
perl Makefile.PL
make
make test
make install
should work. Since this is beta code, you might prefer to install it
in a private library. In that case, replace the first command with
perl Makefile.pl PREFIX=/home/sally/myperl
or whatever.
Sanford Morton