CODEWITHSUNDEEP
pythonlistdictionary
Edouard
Edouard

Reputation: 149

Transform dictionaries to a list of dictionaries according to their values

I have different dictionaries that I want to regroup into a list of dictionaries, taking into account their different values:

Dictionaries :

[{'language': 'de', 'suggestion': 'fressen', 'comment': 'for animals'},
{'language': 'de', 'suggestion': 'essen', 'comment': ''},
{'language': 'fr', 'suggestion': 'manger', 'comment': ''},
{'language': 'fr', 'suggestion': 'bouffer', 'comment': 'slang'}]

List of dictionaries:

[{'language': 'de', 'suggestion': ['fressen', 'essen'], 'comment': ['for animals', '']}, \
{'language': 'fr', 'suggestion': ['manger', 'bouffer'], 'comment': ['', 'slang'}]

I am still a python beginner, and I don't know where to start, to group all the dictionaries with the same 'language': 'de' to the same dict.

Upvotes: 2

Views: 279

Answers (4)

utdemir
utdemir

Reputation: 27216

A straightforward solution, assuming all dicts's keys are same:

ld = [{'language': 'de', 'suggestion': 'fressen', 'comment': 'for animals'},
{'language': 'de', 'suggestion': 'essen', 'comment': ''},
{'language': 'fr', 'suggestion': 'manger', 'comment': ''},
{'language': 'fr', 'suggestion': 'bouffer', 'comment': 'slang'}]

langs = {i['language'] for i in ld}

d = []

for lang in langs:
    d.append({"language": lang})
    for key in ld[0].keys() - ["language"]:
        d[-1][key] = [i[key] for i in ld if i["language"] == lang]

print(d)

Outputs: 

[{'comment': ['', 'slang'],
  'language': 'fr',
  'suggestion': ['manger', 'bouffer']},
 {'comment': ['for animals', ''],
  'language': 'de',
  'suggestion': ['fressen', 'essen']}]

If you choose to use @FMc's data structure described on his comment on your question, you can use that long one-liner, does everything with nested comprehensions:

d = {lang: 
        {i["suggestion"]: i["comment"]
        for i in ld if i["language"] == lang} 
    for lang in {i["language"] for i in ld}}

print(d)

Outputs:

{'de': {'essen': '', 'fressen': 'for animals'},
 'fr': {'bouffer': 'slang', 'manger': ''}}

Upvotes: 4

Artsiom Rudzenka
Artsiom Rudzenka

Reputation: 29111

Not very pythonic but it also works:

data = [{'language': 'de', 'suggestion': 'fressen', 'comment': 'for animals'},
{'language': 'de', 'suggestion': 'essen', 'comment': ''},
{'language': 'fr', 'suggestion': 'manger', 'comment': ''},
{'language': 'fr', 'suggestion': 'bouffer', 'comment': 'slang'}]

def prepare(inputData):
    outputData = []
    if inputData and isinstance(inputData[0], dict) and inputData[0].get('language'):
        keys = inputData[0].keys()
        del keys[keys.index('language')]
    else:
        #do something - raise issue or return
        return outputData 
    for l in set(line['language'] for line in inputData):
        langData = {"language": l}
        langData.update([(k, [line[k] for line in inputData if line["language"] == l]) for k in keys])
        outputData.append(langData)
    return outputData

prepare(data)

Upvotes: 1

cwb
cwb

Reputation: 925

Using reduce gives you clean separation of the merge and the merge logic:

#!/usr/bin/env python

from pprint import pprint
from collections import defaultdict

def group(grouped, ungrouped):
    group = grouped[ungrouped['language']]
    group['language'] = ungrouped['language']
    group['suggestion'].append(ungrouped['suggestion'])
    group['comment'].append(ungrouped['comment'])
    return grouped

ungrouped = [{'language': 'de', 'suggestion': 'fressen', 'comment': 'for animals'},
             {'language': 'de', 'suggestion': 'essen', 'comment': ''},
             {'language': 'fr', 'suggestion': 'manger', 'comment': ''},
             {'language': 'fr', 'suggestion': 'bouffer', 'comment': 'slang'}]

grouped = reduce(group, ungrouped, defaultdict(lambda: defaultdict(list))).values()

# Convert defaultdict to dict for pretty printing.
pprint([dict(group) for group in grouped])

Upvotes: 1

Roman Bodnarchuk
Roman Bodnarchuk

Reputation: 29727

A bit complicated solution, but nevertheless:

>>> dicts = [{'language': 'de', 'suggestion': 'fressen', 'comment': 'for animals'},
... {'language': 'de', 'suggestion': 'essen', 'comment': ''},
... {'language': 'fr', 'suggestion': 'manger', 'comment': ''},
... {'language': 'fr', 'suggestion': 'bouffer', 'comment': 'slang'}]
>>> main_key = 'language'
>>> result_dict = dict()
>>> for d in dicts:
...     for key, value in d.iteritems():
...             if key == main_key:
...                     result_dict.setdefault(d[main_key], dict())[main_key] = value
...             else:
...                     result_dict.setdefault(d[main_key], dict()).setdefault(key, list()).append(value)
...
>>> result_dict.values()
[{'comment': ['', 'slang'], 'language': 'fr', 'suggestion': ['manger', 'bouffer']}, {'comment': ['for animals', ''], 'language': 'de', 'suggestion': ['fressen', 'essen']}]

Basing on main_key value we join all the others dictionaries in the list on this key.

Upvotes: 1

Related Questions