can't plot predict line in R

Question

I'm using those data :

'data.frame':   1584 obs. of  3 variables:
$ Individual: Factor w/ 3 levels "AG201","AG202",..: 1 1 1 1 1 1 1 1 1 1 ...
$ Used      : Factor w/ 2 levels "no","yes": 2 2 2 2 2 2 2 2 2 2 ...
$ NDVI      : int  4724 4576 4894 4297 4670 4932 4346 3810 3481 4058 ...

I'm doing a glm with "NDVI" as a continuous explanatory variable, and then I'm plotting the model through the scatterplot of the data (I'm reproducing the same script as in the R book, Crawley, p.596)

model<-glm(Used~NDVI,binomial);
xv<-seq(0,10000,0.2);
yv<-predict(model,list(NDVI=xv),type="response");
plot(NDVI,Used);
lines(xv,yv);

My problem is that no line appears on my graph...

Any idea what's wrong?

Answer 1

Following Gavin's insight, here's a suggestion:

plot(NDVI, as.numeric(Used)-1 )
lines(xv,yv)

Factors are integer vectors starting at 1L with assignments by default in alpha order of the labels. So you should be OK with "no" < "yes" leading to the No's being 1 and hte Yes's being 2 and then shifting down to the correct scale [0,1]. You may need to also look at str(yv)