SciPy root inconsistent values between function calls

Question

Using opt.root, I want to store the arguments with which the function is called.

import scipy.optimize as opt

x0=[0, 0]
prev_x=x0
count=0
def fun(x):
    global prev_x,count
    print('---')
    print('iteration',count)
    count+=1
    print('last iteration',prev_x)
    print('this iteration',x)
    prev_x=x
    return x[0]**2*x[1]+1,x[1]-x[0]+2

result=opt.root(fun,x0,method='lm')

I would expect that for every iteration step, the values saved in last_x should match the current value of the previous iteration. Running the above code, I get different values between e.g. iterations 8 and 9.

Am I missing something fundamental about the behavior of Python and/or SciPy, or is this a bug?

[edit] I shortened the question to the essential point.

Answer 1

assigning lists does not copy values across, just gives it a different name, so changes to the list through one name are reflected in the other. To get the behavior you expect, change the relevant line to

    prev_x=x.copy()

(note the copy() bit)