CODEWITHSUNDEEP
pythonscipy
crevell
crevell

Reputation: 460

scipy.optimize.curve_fit TypeError

In attempting to fit an exponential to data, scipy.optimize.curve_fit returns a TypeError

x = np.array([0.,1200.02220551,3600.06661654,6000.11102756,8400.15543858,10800.19984961])
y = np.array([0.51057636,0.63187347,0.72030091,0.75168574,0.79036657,0.81551974])

def f(x,p1,p2):                                                                  
    p1*np.exp(x*p2)

popt, pcov = curve_fit(f, x, y)

Unfortunately this returns the following error:

TypeError: unsupported operand type(s) for -: 'NoneType' and 'float'

Any suggestions of what has gone wrong here would be greatly appreciated.

Upvotes: 0

Views: 856

Answers (1)

MB-F
MB-F

Reputation: 23647

f needs to return its result.

def f(x,p1,p2):                                                                  
    return p1*np.exp(x*p2)

Note: Don't be surprised if you get a different error (optimal parameters not found). The data looks logarithmic rather than exponential. Switching x and y (curve_fit(f, y, x)) might work with an exponential function.

Upvotes: 1

Related Questions