Problem and solution: Patch for new "stats" command

[Philipp K. J. <ja...@ie...>]

Here is a recurrent problem I often have:

Frequently, I would like to plot some data after
subtracting off the mean. Or I would like to 
normalize the variation by dividing with the 
standard deviation. Or I'd like to form a normalized
histogram, by passing the number of records in 
the data set to "smooth frequency".

Currently, I always have to invoke an external
program to find these quantities, which is always
a little inconvenient in interactive use, and pretty
painful when scripting gnuplot. When not on Unix,
it may be quite difficult!

For the last few weeks,  Zoltán and I have been 
working on a command that calculates the most
important such quantities from a data file, displays
them, and (optionally) assigns them to variables in
the current gnuplot session. 

You can see some examples of what you can do 
with this command here:
        http://www.phyast.pitt.edu/~zov1/gnuplot/patch/stats.html
and you can read the full documentation here:
        http://www.phyast.pitt.edu/~zov1/gnuplot/patch/stats_help.html

I uploaded a patch with our changes to sourceforge.

We'd like to hear feedback and suggestions. Is this
useful? Are we missing anything?

We'd also like to encourage everyone to build the
patch and play with it - each additional user finds
a new class of bugs!

Best,

                Ph.