data = [{u'class': u'A'}, {u'class': u'A'}, {u'class': u'B'}]
I want to get a list with all existing class values in data. Is there any efficient way to sort out double values?
output = [{u'class': u'A'}, {u'class': u'B'}]
Asked
Active
Viewed 112 times
-2
flowerflower
- 327
- 1
- 3
- 10
-
Are the keys for all the dicts always the same? – Psidom Aug 01 '16 at 15:12
-
Do you mean "efficient" in small execution time for large `data` or in small memory usage or in short code or other? – Rory Daulton Aug 01 '16 at 15:17
1 Answers
0
Use a set instead of a list:
output = set(d['class'] for d in data)
RemcoGerlich
- 30,470
- 6
- 61
- 79
-
-
This gives a `KeyError`. Did you want to use `u'class'` rather htan `'value'`? And as someone else has pointed out, the result is not what is wanted. – Rory Daulton Aug 01 '16 at 15:19
-
-
And thank you Poke for linking, I didn't know that I should look for the words dublicate and dict. With `list(set(i['class'] for i in data))` I get `[u'A, u'B]`. And with `[dict(t) for t in set([tuple(d.items()) for d in data])]` I get `[{u'class': u'B'}, {u'class': u'A'}]` . – flowerflower Aug 01 '16 at 21:50
-
@flowerflower: you can also turn the dicts into frozendicts (same but immutable) as they can be placed into sets: `set(frozendict(d) for d in data)`. Then put `list()` around that if you want to turn it into a list again. Order isn't preserved though. – RemcoGerlich Aug 02 '16 at 07:00