fitting exponential decay with no initial guessing

Question

Does anyone know a scipy/numpy module which will allow to fit exponential decay to data?

Google search returned a few blog posts, for example - http://exnumerus.blo

Answer 1

I would use the scipy.optimize.curve_fit function. The doc string for it even has an example of fitting an exponential decay in it which I'll copy here:

>>> import numpy as np
>>> from scipy.optimize import curve_fit
>>> def func(x, a, b, c):
...     return a*np.exp(-b*x) + c

>>> x = np.linspace(0,4,50)
>>> y = func(x, 2.5, 1.3, 0.5)
>>> yn = y + 0.2*np.random.normal(size=len(x))

>>> popt, pcov = curve_fit(func, x, yn)

The fitted parameters will vary because of the random noise added in, but I got 2.47990495, 1.40709306, 0.53753635 as a, b, and c so that's not so bad with the noise in there. If I fit to y instead of yn I get the exact a, b, and c values.