[R] Outliers in Binary Logistic Regressions

[arun]

Hi,
Not sure how your data looks like.  With the sample data below, the code works.

Try this:
set.seed(1)
dat1<-data.frame(MIGRATION=sample(c(0,1),100,replace=TRUE),distance=sample(40:80,100,replace=TRUE))
RR.rebuild<-glm(MIGRATION~distance,data=dat1,subset=!(1:100 %in% c(56,23,20,9,19)),family=binomial(link="logit"))
 RR.rebuild

#Call:  glm(formula = MIGRATION ~ distance, family = binomial(link = "logit"), 
  #  data = dat1, subset = !(1:100 %in% c(56, 23, 20, 9, 19)))

#Coefficients:
#(Intercept)     distance  
 #-0.0781611   -0.0004483  

#Degrees of Freedom: 94 Total (i.e. Null);  93 Residual
#Null Deviance:        131.4 
#Residual Deviance: 131.4     AIC: 135.4 
A.K.




----- Original Message -----
From: Marcus Tullius <tullius at europe.com>
To: r-help at r-project.org
Cc: 
Sent: Wednesday, September 5, 2012 3:42 PM
Subject: [R] Outliers in Binary Logistic Regressions

Hallo folks,

I know I should not ask the same question again. But I have a problem I cannot solve and without the solution I am stuck and lost, unable to get along with my work!

Someone suggested I should try the code below in order to eliminate the outliers from my data. I did as I was told, but I got a negative reply. The code did not function. I am including it here so that, if possible, someone may correct it for me. That would really be very much appreciated!

My data has 1439 rows. 


*RR.rebuild <- glm(RR, subset=remove)
glm(RR, subset=!(1:1439 %in% c(56,303,365,391,512,746,859,940,1037,1042,1138,1355))
influence(RR.rebuild) 
influence.measures(RR.rebuild)*

Many thanks in advance for any help and sorry for being annoyingly persistent!
Francisco

    [[alternative HTML version deleted]]

______________________________________________
R-help at r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.