Does S-Plus have a built-in graphics parameter or plotting argument
primitive that sizes the points on the basis of their "influence"? In
other words, the size of the plot symbol represents the extent of influence
each point exerts on the Pearson correlation coefficient. (The influence of
a point is the amount the correlation would change if that point were deleted.)
I've seen this in several other pieces of statistics software (JMP, Systat)
but I can't seem to run it down in S-Plus. It may be simply a matter of
terminology, or it may not be implemented. I've run through all my sources
here at home, but can't seem to find anything that corresponds.
I can think of a couple of different ways this might be implemented (e.g. a
function that generates the variable fed to the "size" parameter), but
don't want to waste any time doing something that may be right in front of
my nose, or easily available elsewhere. I tried searching the S-News
archives but the server has been down and/or inaccessible from my location
for nearly two days now.
Thanks.
Dr. Marc R. Feldesman
email: feldesmanm@pdx.edu
email: feldesman@ibm.net
fax: 503-725-3905
"Math is hard. Let's go to the mall" Barbie
Powered by: Monstrochoerus - the 300 MHz Pentium II
-----------------------------------------------------------------------
This message was distributed by s-news@wubios.wustl.edu. To unsubscribe
send e-mail to s-news-request@wubios.wustl.edu with the BODY of the
message: unsubscribe s-news
|