confusing error when using minimize of lmfit in python - python-3.5

Question

my code is as following:
import numpy as np
from math import *
from scipy.optimize import *
import scipy.optimize as opt
from lmfit import Minimizer, Parameters, report_fit
import lmfit as lf
f = open('data.txt','r')
lines=f.readlines()
n1=[]
n2=[]
n=[]
h=[]
for x in lines:
x=x.strip() # remove \n before splitting the line
n1.append(x.split('\t')[0])
n2.append(x.split('\t')[1])
n.append(x.split('\t')[2])
h.append(x.split('\t')[3])
f.close()
n1 = [float(i) for i in n1]
n2 = [float(i) for i in n2]
n = [float(i) for i in n]
h = [float(i) for i in h]
# convert a list into an array
n1 = np.array(n1)
n2 = np.array(n2)
n = np.array(n)
h = np.array(h)
def fith(params,n1,n2,n,h):
a1 = params['p1']
b1 = params['p2']
a2 = params['p3']
b2 = params['p4']
model = (a1 + b1*n) * n1 + (a2 + b2*n) * n2
return model - h
params = Parameters()
params.add('p1',value=1.0)
params.add('p2',value=1.0)
params.add('p3',value=1.0)
params.add('p4',value=1.0)
out = minimize(fith,params,args=(n1,n2,n,h))
print(out)
after run, I got the error as following:
#
Traceback (most recent call last):
File "E:\new model\calculate_H_v2.py", line 50, in
out = minimize(fith,params,args=(n1,n2,n,h))
File "E:\softwares\python\lib\site-packages\scipy\optimize_minimize.py", line 481, in minimize
return _minimize_bfgs(fun, x0, args, jac, callback, **options)
File "E:\softwares\python\lib\site-packages\scipy\optimize\optimize.py", line 943, in _minimize_bfgs
gfk = myfprime(x0)
File "E:\softwares\python\lib\site-packages\scipy\optimize\optimize.py", line 292, in function_wrapper
return function((wrapper_args + args))
File "E:\softwares\python\lib\site-packages\scipy\optimize\optimize.py", line 703, in approx_fprime
return _approx_fprime_helper(xk, f, epsilon, args=args)
File "E:\softwares\python\lib\site-packages\scipy\optimize\optimize.py", line 637, in _approx_fprime_helper
f0 = f(((xk,) + args))
File "E:\softwares\python\lib\site-packages\scipy\optimize\optimize.py", line 292, in function_wrapper
return function(*(wrapper_args + args))
File "E:\new model\calculate_H_v2.py", line 35, in fith
a1 = params['p1']
IndexError: only integers, slices (:), ellipsis (...), numpy.newaxis (None) and integer or boolean arrays are valid indices
#
I could not figure it out why I got this kind of error after read through the questions. Can you help me out?
Thanks in advance.
Jing

Answer 1

I think the basic problem is that you are using scipy.optimize.minimize() instead of lmfit.minimize(). That is, you import * from scipy.optimize, then you import Minimizer from lmfit.
That is, using from lmfit import minimize, Parameters, report_fit, or use
mini = Minimizer(fith,params,args=(n1,n2,n,h))
out = mini.minimize()
should make your script use lmfit.minimize() which looks to me like it should then work.
This is why import * is painful: you have a hard time telling where symbols came from.
As a side comment, you can probably use numpy.loadtxt() to simplify the reading of your data to numpy arrays.