Append list to a python list

Question

I am trying to run a random walk on a two-dimensional grid with four equiprobable actions [right(1), left(-1), upward(1), downward(-1)]. As the random walker moves forward, I want to store its location (x, y coordinates) in the list totSteps. The x,y coordinates (that is, current location) will be updated on every step as a variable curr_loc. As you can see in the printed output, the first list (curr_loc) is the updated current location, and the second list (totSteps) supposedly contains the steps taken so far. There were 5 steps taken, and as such we have 10 outputs. Every time I append the curr_loc to totSteps; all the previous coordinates get replaced with the current one. What's the reason?

steps = [1,-1,1,-1]
totSteps = [] # stores all the 5 steps taken but doesn't work 
# random walk with four steps left, right, 
# upward, downward on two dimensional grid
curr_loc = [0,0]
N = 5
for i in range(N):
 ranNums = np.random.randint(0,4) # picks one of four actions 
 if ranNums == 0 or ranNums == 1: # change x-coordinate
 curr_loc[0] += steps[ranNums] # taking the step 
 print(curr_loc) # current location of random walker 
 totSteps.append(curr_loc)
 print(totSteps) # append current location of random walker 
 elif ranNums == 2 or ranNums == 3: # chanfe y-coordinate 
 curr_loc[1] += steps[ranNums]
 print(curr_loc)
 totSteps.append(curr_loc) 
 print(totSteps)

The output of the code is given below:

>[1, 0] # curr_loc
>[[1, 0]] # totSteps
>[1, -1]
>[[1, -1], [1, -1]]
>[1, 0]
>[[1, 0], [1, 0], [1, 0]]
>[1, -1]
>[[1, -1], [1, -1], [1, -1], [1, -1]]
>[0, -1]
>[[0, -1], [0, -1], [0, -1], [0, -1], [0, -1]]

• Andrej Kesely: You're appending the same curr_loc to the list every iteration. Try totSteps.append(curr_loc[:])
• Muhammad Bilal Shahid: Thanks. It works. When I print curr_loc in the iteration, it prints out the updated value. I still don't understand why all previous values get replaced with the current value. Python append method appends the value (or list) to the end of list.
• Andrej Kesely: @fireshadow52's answer is good at explaining it :) Basically, you're adding the same one reference to the output list, so any change to any of the list affects all other lists.

Answer 1

To further expand on @AndrejKesely's answer, you're not defining a new list object when a new loop iteration starts, and so whenever you change the values in curr_loc, since you're essentially appending another reference to curr_loc to totSteps, you have five references to the same object and that's why you're getting the same values.

Andrej's solution of curr_loc[:] means that you're effectively making a copy of the entire list and storing that rather than a reference to curr_loc.

• Muhammad Bilal Shahid: Thanks for your explanation. I had never faced this issue before. I don't have much experience, but I think the concept of reference is common in C++. I will look into it further.
• fireshadow52: No worries! Yes, "pass-by-value" and "pass-by-reference" are common in C++ as well, but there it's a bit more obvious when a function parameter is a reference parameter.
• Muhammad Bilal Shahid: Yes. I need to look into it closely. I spent so much time figuring out the issue.