2

The given code below is a part of a code i am working on. Here x is a numpy array and f is an array of arrays as given in the example below the code. i have to update all values of the array x and thus i use run a loop to do so. I am using an iterative scheme and the values are updated using the last known value of x. However the loop runs fine for the first time, but the next time it takes the value of x that it got when the 1st loop was run. But i intend to take the old value of x for all the loops. Can anyone help me out please?

import numpy as np
def mulsucc(x0,f,n):
    old=x0
    for k in range(1,100):
        for i in range(0,n):
            print "old" + str(old)
            x0[i]=np.sum(old*(f[i][0:len(f[i])-1])) + 2
            print "new" + str(x0)
        return x0
x0=np.array([0,0])
f=np.array([[4,5,-20],[2,7,-10]])
print mulsucc(x0,f,2)

the output comes out to be:

old[0 0]
new[2 0]
old[2 0]
new[2 6]
[2 6]

however I want that the second old value should also be [0 0].

Mazdak
  • 105,000
  • 18
  • 159
  • 188
Shivam Saxena
  • 83
  • 7
  • The `k` loop is not "used" since you do a return after the `i` loop. Is that on purpose? – toine Nov 05 '15 at 14:49
  • It is a part of a bigger code.. so actually i forgot to remove the k loop.. the problem i face is currently only concerned with the i loop.. – Shivam Saxena Nov 05 '15 at 14:53

1 Answers1

1

you should change:

old = x0

to:

old = x0.copy()

if you want to keep old unchanged. The problem here is that old and x0 are pointers to the same object, so when you alter x0 with the line:

x0[i]=np.sum(old*(f[i][0:len(f[i])-1])) + 2

you are also altering old. old is x0 will be True. There is more to it if you want to dig a bit into related topic.

Community
  • 1
  • 1
toine
  • 1,946
  • 18
  • 24