regressions and correlation



Justin Harford wrote:
> Basically the moral of the story is that regressions and correlations  
> are useful in chem.  I could talk about correlations but that would  
> prolong this message.  Is there a way to do either of these with  
> maxima?  Sure would be nice if I didn't have to open up my vm to do  
> the linest function in excel under win.
>
>   
I think you would do better to look at R or Dataplot.   I haven't used R 
but have heard good things about it.  I have used Dataplot and it can do 
everything; but it's command line oriented; last time I looked (about 5 
years ago) the tcl/tk interface was not robust.   I just looked and 
apparently they have improved the interface.  The underlying base is 
more than 30 years old and does (almost) everything.  I guess it was 
work for interns/post docs for all those years.  The people responsible 
at NIST have done a good job.

http://www.itl.nist.gov/div898/software/dataplot/

Unless your an expert
http://www.itl.nist.gov/div898/software/dataplot/handbook.htm
is very good.

R
http://openwetware.org/wiki/R_Statistics
http://socserv.mcmaster.ca/jfox/Misc/Rcmdr/

To put it politely, I think Excel is limited for statistics analysis.

RayR