s-news
[Top] [All Lists]

cor.test and logistic regression

To: s-news@lists.biostat.wustl.edu
Subject: cor.test and logistic regression
From: Thierry de Meeûs <Thierry.DeMeeus@cepm.mpl.ird.fr>
Date: Fri, 16 Feb 2001 18:12:47 +0100
Dear S-Plus users

I am trying to compute and test the significance of spearman correlation coefficient out of a big data set and cannot figure out how to do this. I do not know how to make S-Plus to understand that the two variables named "var1" and "var2" I desire to compare are in the file "myfile". All my attempts with "cor.test" has led to an error message.

I have also a problem with a logisitic regression (done by glm). The residuals are dispersed into two well separated sets of points. It would be very useful to select each of these two sets in order study them independently. I however did not find the procedure in S+ manuals that would allow to do so. Any advice would be useful.

I work with S-plus 2000 for windows with Windows Me, on a PC Pentium III 500 MHz, 256 Mega of RAM.

Thank you

Sincerely

Thierry

Thierry de Meeûs
Centre d'Etude sur le Polymorphisme des Micro-Organismes
(CEPM), UMR CNRS-IRD 9926
Equipe: "Evolution des Systèmes Symbiotiques"

Centre IRD de Montpellier
911 Avenue Agropolis, B.P. 5045
34032 Montpellier Cedex 1, France.

         /|||||\
       /||||||||\_
     /  ^   ^   ^ \    ___
  __/  ^  ^   ^  ^ \__/___\
/ @  ) °°°°°°°°°°°° \______\
> @  ) ^  ^  ^  ^   /_ ___ /
\___/   ^  ^  ^  ^ /  \___/
     \   ^  ^  ^ _/
      \(||)|||||/
        \||||||/

Tel:         (33) 4 67 41 63 10
Secrétariat: (33) 4 67 41 61 97
Fax:         (33) 4 67 41 62 99

<Prev in Thread] Current Thread [Next in Thread>
  • cor.test and logistic regression, Thierry de Meeûs <=