CODEWITHSUNDEEP
modelcurvedata-fittingasymptotedecay
ChrisJ
ChrisJ

Reputation: 1

R model fit for a decay curve

I am trying to get a decay curve fitted to this data to get the values for other distances in R.

data

I have tried a few self-starting models (SSmicem, SSsymp), but these dont seem to appreciate that I want this to be a decay curve and I end up with relatively large values towards the tail end here.

I am wishing that the model comes to the asymptote of close to 0, but all models i have produced are giving large values at the tail end (larger than the observed values). Is there a way i can set this asymptote?

Upvotes: 0

Views: 41

Answers (1)

JJacquelin
JJacquelin

Reputation: 1705

Did you draw the points in order to observe the shape of an expected curve ? Drawing it in log-log coordinates makes the inspection easier.

By inspection one see that there is only one point in the range 0.03<y<0.5 which is a wide range (Thus very badly defined ). As a consequence one cannot expect a reliable fitting.

Certainly they are many kind of functions which could be fitted. Without clue about the kind of function expected from physical consideration a chosen function will have no physical signifiance. For example the function below is probably without practical interest even with a not too bad fitting. enter image description here

Upvotes: 0

Related Questions