Another alternative is to use SSlogis which is very similar to the
model you're fitting except with one additional parameter:
Asym <- 3
xmid <- 0
scal <- 10
model <- nls(bod ~ SSlogis(time, Asym, xmid, scal), data = mydata)
summary(model)
plot(bod ~ time, mydata)
newdata <- data.frame(time = seq(1
Hi, Michael,
I think the SPSS answer is wrong. Your starting values are way off.
Look at this plot for verification:
con <- textConnection("time bod
11 0.47
22 0.74
33 1.17
44 1.42
55 1.60
67 1.84
79 2.19
8 11 2.17")
mydata <- read.table(con, header = TRUE)
close(co
Hi All,
I'm trying to run nls on the data from the study by Marske (Biochemical
Oxygen Demand Interpretation Using Sum of Squares Surface. M.S. thesis,
University of Wisconsin, Madison, 1967) and was reported in Bates and Watts
(1988).
Data is as follows, (stored as mydata)
time bod
11 0.
3 matches
Mail list logo
