convert Python list to ordered unique values
Question:
I encounter many tasks in which I need to filter python (2.7) list to keep only ordered unique values. My usual approach is by using odereddict
from collections:
from collections import OrderedDict
ls = [1,2,3,4,1,23,4,12,3,41]
ls = OrderedDict(zip(ls,['']*len(ls))).keys()
print ls
the output is:
[1, 2, 3, 4, 23, 12, 41]
is there any other state of the art method to do it in Python?
- Note – the input and the output should be given as
list
edit – a comparison of the methods can be found here:
https://www.peterbe.com/plog/uniqifiers-benchmark
the best solution meanwhile is:
def get_unique(seq):
seen = set()
seen_add = seen.add
return [x for x in seq if not (x in seen or seen_add(x))]
Answers:
If you need to preserve the order and get rid of the duplicates, you can do it like:
ls = [1, 2, 3, 4, 1, 23, 4, 12, 3, 41]
lookup = set() # a temporary lookup set
ls = [x for x in ls if x not in lookup and lookup.add(x) is None]
# [1, 2, 3, 4, 23, 12, 41]
This should be considerably faster than your approach.
You could use a set like this:
newls = []
seen = set()
for elem in ls:
if not elem in seen:
newls.append(elem)
seen.add(elem)
Define a function to do so:
def uniques(l):
retl = []
for x in l:
if x not in retl:
retl.append(x)
return retl
ls = [1,2,3,4,1,23,4,12,3,41]
uniques(ls)
[1, 2, 3, 4, 23, 12, 41]
Another solution would be using list comprehension like this:
[x for i, x in enumerate(ls) if x not in ls[:i]]
Output:
[1, 2, 3, 4, 23, 12, 41]
I encounter many tasks in which I need to filter python (2.7) list to keep only ordered unique values. My usual approach is by using odereddict
from collections:
from collections import OrderedDict
ls = [1,2,3,4,1,23,4,12,3,41]
ls = OrderedDict(zip(ls,['']*len(ls))).keys()
print ls
the output is:
[1, 2, 3, 4, 23, 12, 41]
is there any other state of the art method to do it in Python?
- Note – the input and the output should be given as
list
edit – a comparison of the methods can be found here:
https://www.peterbe.com/plog/uniqifiers-benchmark
the best solution meanwhile is:
def get_unique(seq):
seen = set()
seen_add = seen.add
return [x for x in seq if not (x in seen or seen_add(x))]
If you need to preserve the order and get rid of the duplicates, you can do it like:
ls = [1, 2, 3, 4, 1, 23, 4, 12, 3, 41]
lookup = set() # a temporary lookup set
ls = [x for x in ls if x not in lookup and lookup.add(x) is None]
# [1, 2, 3, 4, 23, 12, 41]
This should be considerably faster than your approach.
You could use a set like this:
newls = []
seen = set()
for elem in ls:
if not elem in seen:
newls.append(elem)
seen.add(elem)
Define a function to do so:
def uniques(l):
retl = []
for x in l:
if x not in retl:
retl.append(x)
return retl
ls = [1,2,3,4,1,23,4,12,3,41]
uniques(ls)
[1, 2, 3, 4, 23, 12, 41]
Another solution would be using list comprehension like this:
[x for i, x in enumerate(ls) if x not in ls[:i]]
Output:
[1, 2, 3, 4, 23, 12, 41]