Question

Sorry if this is a novice question, but I don’t understand it. I am fitting a sigmoid curve to my data with glm(). This works, I can plot the output and I see a nice sigmoid curve.

However, how do I get R to return the final values it has fit? As I understand it, R fits the data to logit(y) = b0 + b1x, but when I do > summary(glm.out) I only get

Call:
glm(formula = e$V2 ~ e$V1, family = binomial(logit), data = e)

Deviance Residuals: 
       1         2         3         4         5         6         7  
-0.00001  -0.06612  -0.15118  -0.34237   0.20874   0.08724  -0.19557  

Coefficients:
            Estimate Std. Error z value Pr(>|z|)
(Intercept)  -24.784     20.509  -1.208    0.227
e$V1           2.073      1.725   1.202    0.229

(Dispersion parameter for binomial family taken to be 1)

    Null deviance: 4.60338  on 6  degrees of freedom
Residual deviance: 0.23388  on 5  degrees of freedom
AIC: 5.8525

Number of Fisher Scoring iterations: 8

How do I get b0 and b1?

A sample dataset:

The Z column can be ignored.

X   Y   Z
0.0 0.0 6
6.5 0.0 3
8.8 0.333333333333  3
10.5    0.2 10
11.1    0.0 3
11.25   0.166666666667  6
12.0    0.2 5
12.75   0.5 6
13.4    0.333333333333  3
13.5    0.2 5
14.25   0.5 6
15.0    0.333333333333  6
15.7    0.666666666667  3
15.75   0.666666666667  6
16.5    0.833333333333  6
17.25   0.555555555556  9
18.0    1.0 3

Answer 1

You get the fitted values via the fitted() method, i.e. fitted(glm.out). However, you want the estimated coefficients not the fitted values, and for that you want the coef() method, as in coef(glm.out).