CODEWITHSUNDEEP
pythoncurve-fitting
Alan
Alan

Reputation: 11

How to create a function of N exponentials in python to fit data

I am trying to write a code in python that fit a sum of exponetials to some data. However I want to have the freedom to choose how many exponetials I use.

def fcn(x,nexp):
    return sum(ai * np.exp(-Ei * x)  for ai, Ei in range(nexp))

I have tried to define the above and then call it inside scipy.optimize.curve_fit but I get the error "'int' object is not iterable". Any help would greatly be appreciated. At the moment I just keep defining new functions and adding terms as follow

def f1(t,a1,E1):
    return a1*np.exp(-E1*t)


def f2(t,a1,E1,a2,E2):
    return a1*np.exp(-E1*t)+a2*np.exp(-E2*t)


def f3(t,a1,E1,a2,E2,a3,E3):
    return a1*np.exp(-E1*t)+a2*np.exp(-E2*t)+a3*np.exp(-E3*t)


def f4(t,a1,E1,a2,E2,a3,E3,a4,E4):
    return a1*np.exp(-E1*t)+a2*np.exp(-E2*t)+a3*np.exp(-E3*t)+a4*np.exp(-E3*t)

Upvotes: 1

Views: 98

Answers (1)

Jason Stein
Jason Stein

Reputation: 724

Instead of passing each a1, E1 as separate arguments, you could have one function that accepts a list of two-tuples (e.g. [(a1,E1),(a2,E2)])then iterate through.

def f(t, lst): sum = 0 for a, E in lst: sum += a*np.exp(-E*t) return sum This could be written shorter, I gave it step-by-step for clarity.

Upvotes: 3

Related Questions