Python and collections.counter - how to make elements dynamic

Question

I need to add items to a counter and needs to be more dynamic.

  hash_data = [{'campaign_id': 'cid2504649263',
  'country': 'AU',
  'impressions': 9000,
  'region': 'Cairns',
  'utcdt': datetime.datetime(2013, 6, 4, 6, 0)},
 {'campaign_id': 'cid2504649263',
  'country': 'AU',
  'impressions': 3000,
  'region': 'Cairns',
  'utcdt': datetime.datetime(2013, 6, 4, 6, 0)},
 {'campaign_id': 'cid2504649263',
  'country': 'AU',
  'impressions': 3000,
  'utcdt': datetime.datetime(2013, 6, 4, 7, 0)}]

For example, then last element in the list of hashes does not contain region. Yet the below is how I add elements and will get an error.

C = Counter()
for item in hash_data:
    C[item['utcdt'],item['campaign_id'], item['country'], item['region']] += item[metric]

Ideally something like this but of course not work

C = Counter()
for item in hash_data:
    m1 = item.keys()
    m2 = []
    for i in ml:
        if i!='impression':
           ms.add(i)
    C[ml] += item[metric]

What default value would you assume for region if it does not exist? Why don't you use `item.get('region', default)`? — mg007, Jun 12 '13 at 09:07
@mg007 you should add `item.get('region','default')` as an answer. — Burhan Khalid, Jun 12 '13 at 09:10
It need to be dynainc..nit just region....what if I have 100 keys? — Tampa, Jun 12 '13 at 09:10
So the key in your Counter is the intersection of your common fields? — Vlad, Jun 12 '13 at 09:43

score 1 · Answer 1 · answered Jun 12 '13 at 09:46

1

C = Counter()
for item in hash_data:
  C[tuple(item.values())] += item[metric]

answered Jun 12 '13 at 09:46

Vlad

18,195
4
41
71

Not a good idea, because two dictionaries may have different key subsets but the same values – Eric Dec 21 '13 at 23:44

score 0 · Answer 2 · edited May 23 '17 at 11:57

As I mentioned in the comments, you can use the dict.get with a predefined default value which will be returned in case the key is not present in the dict.

default_region = 'default_region'
...
...
item.get('region', default_region)

If you have many keys which you think might be absent from your dict, you can use collections.defaultdict if that suits your purpose (See this discussion). Alternatively, you can simply create a dict of default values containing all possible keys and their corresponding default values. A small example could be:

defaults = {
            'campaign_id': 'default_campaign_id',
            'country': 'default_country',
            'impressions': -1,
            'region': 'default_region',
            'utcdt': datetime.datetime.min,
           }

for item in hash_data:
    C[tuple(item.get(k, dv) for k, dv in defaults.items() if k != metric)] += item.get(metric, defaults[metric])

HTH!

Python and collections.counter - how to make elements dynamic

2 Answers2