Menu
I'm trying to fit a curve over (the tail of) the following data:
[1] 1 1 1 1 1 1 2 1 2 2 3 2 1 1 4 3 2 11 6 2 16 7 17 36 [25] 27 39 41 33 42 66 92 138 189 249 665 224 309 247 641 777 671 532 749 506 315 292 281 130 [49] 137 91 40 27 34 19 1
I'm using the following function in R to accomplish this:
nls(y~a*x*exp(-b*x^2),start=list(a=1,b=1),trace=TRUE)
However, I'm getting the following error:
3650202 : 1 1
Error in numericDeriv(form[[3L]], names(ind), env) : Missing value or an infinity produced when evaluating the model
When using the following, artificial values for x and y, everything works just fine:
y=x*exp(-.5*x^2)+rnorm(length(x),0,0.1)
x [1] 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 [20] 0.95 1.00 1.05 1.10 1.15 1.20 1.25 1.30 1.35 1.40 1.45 1.50 1.55 1.60 1.65 1.70 1.75 1.80 1.85 [39] 1.90 1.95 2.00 2.05 2.10 2.15 2.20 2.25 2.30 2.35 2.40 2.45 2.50 2.55 2.60 2.65 2.70 2.75 2.80 [58] 2.85 2.90 2.95 3.00 3.05 3.10 3.15 3.20 3.25 3.30 3.35 3.40 3.45 3.50 3.55 3.60 3.65 3.70 3.75 [77] 3.80 3.85 3.90 3.95 4.00 4.05 4.10 4.15 4.20 4.25 4.30 4.35 4.40 4.45 4.50 4.55 4.60 4.65 4.70 [96] 4.75 4.80 4.85 4.90 4.95 5.00 y [1] -0.080214106 0.075247488 0.076355116 -0.020087646 0.181314038 0.075832658 0.248303254 [8] 0.364244010 0.453655908 0.347854869 0.514373164 0.384051249 0.618584696 0.515684390 [15] 0.534737770 0.609279111 0.618936091 0.534443863 0.739118585 0.677679546 0.526011452 [22] 0.645645150 0.578274968 0.589619834 0.476186241 0.621638333 0.601663144 0.535981735 [29] 0.518434367 0.581735107 0.423872948 0.445335110 0.340884242 0.317121065 0.342683141 [36] 0.278351104 0.402947372 0.429483276 0.276655872 0.108164828 0.389994138 0.372300257 [43] -0.057320612 0.131271986 0.226212869 0.131171973 0.245970674 0.009926555 0.173465207 [50] 0.141220590 0.280616078 0.108515613 0.117697407 0.130700771 0.058540888 0.251613512 [57] 0.168094899 -0.058382571 0.123306762 -0.048605186 -0.010131767 0.076701962 -0.051982924 [64] 0.058427540 0.144665070 0.063998841 -0.010495697 0.119868854 0.114447318 0.006759691 [71] 0.025041761 -0.178145771 0.041547126 0.122084819 0.034283141 0.209140060 0.197024853 [78] -0.005491966 -0.033260219 -0.028123314 -0.005775553 -0.040781462 0.090024896 0.116390743 [85] -0.017811031 0.094039200 -0.147064060 -0.057249278 0.211587898 -0.066153592 0.032100332 [92] -0.092756136 -0.125906598 0.136937364 0.046453010 0.002000336 -0.134047101 0.089748847 [99] -0.019355567 -0.042158950 0.149594368
Can anyone point out what I'm doing wrong? Thanks for your help.
784
The nls function uses a relative-offset convergence criterion that compares the numerical imprecision at the current parameter estimates to the residual sum-of-squares. This performs well on data of the form y = f ( x , θ ) + ϵ (with var(eps) > 0 ).
The nls function tries to estimate the parameters for a nonlinear model that minimize the sum of squared residuals. For the default algorithm, the left side of formula is the response to be fitted, and the right side should evaluate to a numeric vector the same length as the response.
Well I found the answer to my problem. The starting values for the real data are completely different from the dummy values: a=500 and b=.1 result in a nice fit. Just thought it might be useful to mention that here.
100
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With
