Python, removing duplicates from a list

Question

I suppose the Pythonic way is to use set, which I just learned about. Before I learned about set, I tried the following method which didn't work. Can someone explain what is wrong with this? I got an index out of range error, but I thought the code would never run past the index range for the list as it is written. Please explain what happened.

#!/usr/bin/python
# Remove dupes from a list
def Uniq_list(x):
 for elem in range(len(x)):
 if ( elem == 0 ):
 next
 else:
 if (x[elem] == x[(elem - 1)]):
 x.pop(index(x[elem])
 return x
mylist = ['1', '1', '2', '1', '3', '2']
print Uniq_list(mylist)

• wastl: since you are popping elemtens from top of the list it gets shorter while you are iterating over it, hence the out of range error
• thefourtheye: There are so many problems in your code. Can you please fix them?
• Martijn Pieters: next doesn't do anything; it is a no-op as it just references the built-in function.
• Martijn Pieters: Where is the index() function?
• GL2014: @wastl That makes sense, thanks a lot! As for the errors in the code, I was trying to go from memory (I have since scrapped this entire function and used set()), so I apologize for the errors, I am a beginner. I was just trying to get my point across as to the index range issue.

Answer 1

The for loop produces the range() object just once, but in the loop you are removing elements from the list, making it shorter.

This leads to another problem: you'll be skipping elements. If you remove element 1, all elements after it shift up one position; element 2 is now element 1, etc. But your loop never takes this into account, so it'll skip the element now moved from position 2 to 1.

This is compounded by the fact you only use the preceding element to test set membership against; the skipped element is suddenly the value against which another value is being tested. They could well be different but that doesn't mean earlier elements are not going to be equal.

Further issues: You used an unnamed index() function in your code which means we cannot verify what it does. If it works like the list.index() method you'd be removing the first occurrence of the value from the list object. This could work, but you already had elem as an index into the list, why search for it again?

The next entry on it's own line is just a reference to the next() function, without ever calling it. As such, it is a no-op. You probably meant to use the continue statement here instead.

A simpler version would have been:

for i in range(len(x), -1, -1):
 if x[i] in x[:i]:
 del x

e.g. starting from the end of the list, if the current element is present in the list before this position, remove it. However, using a set would be far more efficient still.

• GL2014: Thanks very much, I will accept when time limit allows me to do so.
• GL2014: My usage of next was intended as a method to skip testing x[0] against x[0 - 1]. I wanted a noop in the sense that I wanted to jump to the next iteration in the loop if elem evaluates to 0. Is that not what it would be doing?
• Martijn Pieters: @GreggLeventhal: No, that's what the continue statement does, though.
• Martijn Pieters: @GreggLeventhal: it is not needed here, really; the else branch is never entered if the if statement was picked. You could just have done if elem != 0: ... instead.
• GL2014: Darn, I wasn't sure if it was next or continue. Thanks again!