Htykuut

I am running nls models in R on several different datasets, using the self-starting Weibull Growth Curve function, e.g.

MOD <- nls(Response ~ SSweibull(Time, Asym, Drop, lrc, pwr), data = DATA)

With data like this, it works as expected:

GOOD.DATA <- data.frame("Time" = c(1:150), "Response" = c(31.2, 20.0, 44.3, 35.2, 

31.4, 27.5, 24.1, 25.9, 23.3, 21.2, 21.3, 19.8, 18.4, 17.3, 16.3, 16.3,

16.6, 15.9, 15.9, 15.8, 15.1, 15.6, 15.1, 14.5, 14.2, 14.2, 13.7, 14.1,

13.7, 13.4, 13.0, 12.6, 12.3, 12.0, 11.7, 11.4, 11.1, 11.0, 10.8, 10.6,

10.4, 10.1, 11.6, 12.0, 11.9, 11.7, 11.5, 11.2, 11.5, 11.3, 11.1, 10.9,

10.9, 11.4, 11.2, 11.1, 10.9, 10.9, 10.7, 10.7, 10.5, 10.4, 10.4, 10.3,

10.1, 10.0, 9.9, 9.7, 9.6, 9.7, 9.6, 9.5, 9.5, 9.4, 9.3, 9.2, 9.1, 9.0,

8.9, 9.0, 8.9, 8.8, 8.8, 8.7, 8.6, 8.5, 8.4, 8.3, 8.3, 8.2, 8.1, 8.0,

8.0, 8.0, 7.9, 7.9, 7.8, 7.7, 7.6, 7.6, 7.6, 7.6, 7.5, 7.5, 7.5, 7.5,

7.4, 7.4, 7.3, 7.2, 7.2, 7.1, 7.1, 7.0, 7.0, 6.9, 6.9, 6.8, 6.8, 6.7,

6.7, 6.6, 6.6, 6.5, 6.5, 6.4, 6.4, 6.4, 6.3, 6.3, 6.2, 6.2, 6.2, 6.1 

6.1, 6.1, 6.0, 6.0, 5.9, 5.9, 5.9, 5.9, 5.8, 5.8, 5.8, 5.8, 5.8, 5.8,

5.8, 5.7))

But with this data set:

BAD.DATA <- data.frame("Time" = c(1:150), "Response" = c(89.8, 67.0, 

51.4, 41.2, 39.4, 38.5, 34.3, 30.9, 29.9, 34.8, 32.5, 30.1, 28.5, 27.0,

26.2, 24.7, 23.8, 23.6, 22.6, 22.0, 21.3, 20.7, 20.1, 19.6, 19.0, 18.4,

17.9, 17.5, 17.1, 23.1, 22.4, 21.9, 23.8, 23.2, 22.6, 22.0, 21.6, 21.1,

20.6, 20.1, 19.7, 19.3, 19.0, 19.2, 18.8, 18.5, 18.3, 19.5, 19.1, 18.7,

18.5, 18.3, 18.0, 17.7, 17.5, 17.3, 17.0, 16.7, 16.7, 16.9, 16.6, 16.4,

16.1, 15.9, 15.8, 15.6, 15.4, 15.2, 15.0, 14.8, 14.7, 14.5, 14.4, 14.2,

14.0, 13.9, 13.7, 13.6, 15.4, 15.2, 15.1, 15.0, 14.9, 14.7, 14.6, 14.5,

14.4, 14.3, 14.4, 14.2, 14.1, 14.0, 13.8, 13.7, 13.6, 13.5, 13.4, 13.2,

13.3, 13.2, 13.1, 13.0, 12.9, 12.8, 12.7, 12.6, 12.5, 12.5, 12.4, 12.3,

12.2, 12.1, 12.1, 11.9, 12.8, 12.7, 12.6, 12.5, 12.4, 14.2, 14.1, 14.0,

14.1, 14.0, 13.9, 13.8, 13.7, 13.7, 13.6, 13.5, 13.4, 13.3, 13.3, 13.2,

13.1, 13.0, 12.9, 12.9, 12.8, 12.7, 12.6, 12.9, 12.8, 12.7, 12.6, 12.5,

12.5, 12.4, 12.3, 12.2))

I get the error;

Error in nls(y ~ cbind(1, -exp(-exp(lrc) * x^pwr)), data = xy, algorithm = "plinear",

  : step factor 0.000488281 reduced below 'minFactor' of 0.000976562

By including the control argument I am able to change the minFactor for GOOD.DATA:

MOD <- nls(Response ~ SSweibull(Time, Asym, Drop, lrc, pwr), data = GOOD.DATA, 

control = nls.control(minFactor = 1/4096))

But the model was running without errors anyway. With BAD.DATA and several other datasets, including control has no effect and I just get the same error message.

Questions

1. How can I change the minFactor for the BAD.DATA?




2. What's causing the error? (i.e. what is it about the data set that triggers the error?)




3. Will changing the minFactor resolve this error, or is this one of R's obscure error messages and it actually indicates a different issue?

It seems the control option does not work in your case, since the code breaks at getInitial while self-starting, that is, before using your provided control parameters. One way would be to try specifying some starting parameters, instead of the naive self-starting. For nls it is often the case that playing with the initial parameters will make-or-break it, not entirely sure for the specific Weibull case though, but should be the same.

To see that you don't arrive to the actual control, you can try with nls.control(printEval = T) and see that there's no print.