non linear regression

[Marta95]

I have this dataset

devRate <- data.frame( T = c(8, 16, 20, 25, 25, 29, 29, 35), 
 Rate = c(0, 0.27, 0.20, 0.69, 0.76, 1, 0.98, 0))

By running this code:

M2 <- nls(Rate~(a*(T^b)*(1-T))^c, data = Avg_species, start = list(a = 12.236, b = 4.013, c = 0.546), algorithm = "port")

I receive this error

Missing value or an infinity produced when evaluating the model

How to solve it?
Thanks

[startz]

The problem is that 1-T is negative and are effectively taking the square root of a negative number.

[Marta95]

@startz
There is no square root in the function for the elaboration of Rate variable.. am I wrong?

Thanks
Marta

[mafw]

This is correct, you get NaN because (1-T) produces a negative number which you then attempt to exponent using a number less than 1.

Too see this you can simply plug in the numbers, for example (1-8)^0.2 produces NAN.