2
$\begingroup$

I have a sequence of data:

 data = {0.647888, 0.522495, 0.454224, 0.417054, 0.396816, 0.385798, 0.379799, 0.376532, 0.374754, 0.373786, 0.373259, 0.372972}

How can I find asymptote of this sequence using linear regression? Can anybody help me how to start? I just got a clue: y=a+b/c^x. Should I use LinearModelFit?

$\endgroup$
2
$\begingroup$
ClearAll[a, b, c]
data = {0.647888, 0.522495, 0.454224, 0.417054, 0.396816, 0.385798, 
   0.379799, 0.376532, 0.374754, 0.373786, 0.373259, 0.372972};

nlm = NonlinearModelFit[data, a + b/ c^x, {a, b, c}, x];
Normal@nlm

$0.372629\, +0.505569 \,\, 1.8367^{-x}$

Limit[Normal[nlm], x -> Infinity]

0.372629

| improve this answer | |
$\endgroup$
  • $\begingroup$ @hermano9, my pleasure. Thank you for the accept. $\endgroup$ – kglr Jun 12 '16 at 21:59
  • 3
    $\begingroup$ This data set is clearly not from any biology experiment as a plot of the data and the fit shows almost no variability from the predicted model. An approximate 95% confidence interval for the value of the asymptote is found with nlm["ParameterConfidenceIntervals"][[1]] which shows as {0.372629, 0.372629}. $\endgroup$ – JimB Jun 12 '16 at 22:01

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.