[R] Logistic Regression Coding Help

[Craig P O'Connell]

Hello. 

I am beginning to analyze my work and have realized that a simple chi-square 
analysis will not suffice for my research, with one notable reason is that data 
are not discrete.  Since my data fit the assumptions of a logistic regression, 
I am moving forward with this analysis.  With that said, I am a beginner with 
R and would grealty appreciate any help!  

Essentially, the point of my work is to determine if sharks are more sensitive 
to repellents based on certain environmental parameters.  

Therefore an incredibly simplified version of my bullsharkdata .txt file would 
look like this (Note:  (1) = low density/visibility and (2) = high 
density/visibility; and behavior = (1) or avoidance behavior) : 
Obs  Density   Visibility  Behavior   Data 
1      1            1         1           
0.9   
2      2            1         1           
0.1 
3      1            2         
1           0.3 
4      2            2         1           0.8 

Here was my attempt at coding: 
bullsharks <- read.table("bullsharkdata.txt", header=T, as.is=T) 
#"bullsharks" is what I named it 
bullsharks$Obs 
bullsharks$Density 
bullsharks$Visibility 
bullsharks$Behavior 
bullsharks$Data 
#or to view all data 
bullsharks 

library(mlogit) 
bullsharks[1:2,] 
bullsharks$Density<-as.factor(bullsharks$Density) 
mldata<-mlogit.data(bullsharks, varying=NULL, choice="Density", shape="wide") 
mlogit.model <- mlogit(Density~1|Visibility+Behavior, data = mldata, 
reflevel="1") 
summary(mlogit.model) 

However, I get an error at the mlogit.model stage.  Is there something wrong 
with my data or with my code?  

Thank you and any help would be incredibly appreciated! 

        [[alternative HTML version deleted]]


______________________________________________
R-help@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.