R help - Dec 2007 - how to interpolate a plot with a logistic curve

hello,

I have this simple question. This is my dataset

	size
1	57
2	97
3	105
4	123
5	136
6	153
7	173
8	180
9	193
10	202
11	213
12	219
13	224
14	224
15	248
16	367
17	496
18	568
19	618
20	670
21	719
22	774
23	810
24	814
25	823

I plot it with:

plot(generalstats[,1], type="b", xlab="Mesi",
ylab="Numero di
vertici", main="");

and try to interpolate with a linear regression with

abline(lm(generalstats[, 
1 
]~ 
c(1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25)),  
lty=3, col="red");

how to interpolate the data with a logistic curve? I cannot find the -  
I suppose easy - solution..

thank you,
Simone

On Wednesday 05 December 2007, Simone Gabbriellini
wrote:
> hello,
>
> I have this simple question. This is my dataset
>
> 	size
> 1	57
> 2	97
> 3	105
> 4	123
> 5	136
> 6	153
> 7	173
> 8	180
> 9	193
> 10	202
> 11	213
> 12	219
> 13	224
> 14	224
> 15	248
> 16	367
> 17	496
> 18	568
> 19	618
> 20	670
> 21	719
> 22	774
> 23	810
> 24	814
> 25	823
>
> I plot it with:
>
> plot(generalstats[,1], type="b", xlab="Mesi",
ylab="Numero di
> vertici", main="");
>
> and try to interpolate with a linear regression with
>
> abline(lm(generalstats[,
> 1
> ]~
> c(1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25)),
> lty=3, col="red");
>
> how to interpolate the data with a logistic curve? I cannot find the -
> I suppose easy - solution..
>
> thank you,
> Simone
>
try:

glm(formula, data, family=binomial())

require(Design)
lrm()


Cheers,

-- 
Dylan Beaudette
Soil Resource Laboratory
http://casoilresource.lawr.ucdavis.edu/
University of California at Davis
530.754.7341

Dear Simone,

you can use the package 'drc' to fit a logistic model, that is a
non-linear regression
model based on the equation c+(d-c)/(1+exp(b(x-e))), to your data (below named
'simone'):


## Fitting a 4-parameter logistic model (also called Boltzmann model)
simone.m1 <- drm(size~x, data=simone, fct=B.4())

## Plotting the data together with the fitted curve
plot(simone.m1, log="")


The fit is not great due to the bend.


Christian