Combine dictionaries with the same keys value-CodePudding

I have a list of dictionaries like this.

dicts = [
 {"key_a": "2022-01-30",
  "key_b": "54",
  "key_c": "3"},

 {"key_a": "2022-01-30", 
  "key_b": "46", 
  "key_c": "2"},

 {"key_a": "2022-01-04",
  "key_b": "32",
  "key_c": "4"}

 {"key_a": "2022-01-30",
  "key_b": "2",
  "key_c": "10"}
]

# check same value of key_a
# Sum key_b 
# keep the first key_c value from the dictionaries

I want to merge those with same "date". I tried with Panda but I am looking for a more simple way

This is the ouput I am looking for

new_dicts = [
 {"key_a": "2022-01-30",
  "key_b": "102",
  "key_c": "3"},

 {"key_a": "2022-01-04",
  "key_b": "32",
  "key_c": "4"}
]

CodePudding user response：

def get_dicts(dicts):
    dates = {d["key_a"] for d in dicts}
    for date in dates:
        yield {
            "key_a": date,
            "key_b": str(sum(map(int, d["key_b"] for d in dicts if d["key_a"] == date))),
            "key_c": next(d["key_c"] for d in dicts if d["key_a"] == date)
        }

print(list(get_dicts(dicts)))

CodePudding user response：

Here's how I would do it

dates = {}

for dict in dicts:
    curr_date = dict['key_a']
    
    if curr_date not in dates:
        dates[curr_date] = dict
        dates[curr_date]['key_b'] = int(dict['key_b']) # convert key_b to int

    else:
        # add current key_b to running total
        dates[curr_date]['key_b']  = int(dict['key_b'])

new_dicts = list(dates.values())

# convert key_b back to string
for dict in new_dicts:
    dict['key_b'] = str(dict['key_b'])

CodePudding user response：

An alternative method (handles string):

new_dicts = []
existing_dates = []
for d in dicts:
    if d['key_a'] in existing_dates:
        for nd in new_dicts:
            if nd['key_a'] == d['key_a']:
                nd['key_b']  = int(d['key_b'])
    else:
        existing_dates.append(d['key_a'])
        d['key_b'] = int(d['key_b'])
        new_dicts.append(d)    


for d in new_dicts:
    d['key_b'] = str(d['key_b'])
print(new_dicts)

Also, I believe you're missing a comma in the data posted in the answer.

Output:

[{'key_a': '2022-01-30', 'key_b': '102', 'key_c': '3'}, {'key_a': '2022-01-04', 'key_b': '32', 'key_c': '4'}]