merge two dictionary lists if value is the same-CodePudding

I have two lists of dictionaries like the ones below:

a = [
    {'name': 'john', 'exam 1': 'python'},
    {'name': 'chris', 'exam 1': 'java'},
    {'name': 'newman', 'exam 1': 'C'},
    {'name': 'sebast', 'exam 1': 'C#'},
    {'name': 'penier', 'exam 1': 'python'},
    {'name': 'alex', 'exam 1': 'go'},
    {'name': 'steve', 'exam 1': 'C#'}
]

b = [
    {'name': 'john', 'exam 2': 'django'},
    {'name': 'newman', 'exam 2': 'java'},
    {'name': 'sebast', 'exam 2': 'C'},
    {'name': 'chris', 'exam 2': 'C#'}
    {'name': 'penier', 'exam 2': 'python'},
    {'name': 'steve', 'exam 2': 'go'},
    {'name': 'alex', 'exam 2': 'C#'}
]

I would like to merge them in one dictionary like the one below:

c = [
    {'name': 'john', 'exam 1': 'python', 'exam 2': 'django'},
    {'name': 'chris', 'exam 1': 'java', 'exam 2': 'C#'},
    {'name': 'newman', 'exam 1': 'C', 'exam 2': 'java'},
    {'name': 'sebast', 'exam 1': 'C#', 'exam 2': 'C'},
    {'name': 'penier', 'exam 1': 'python', 'exam 2': 'python'},
    {'name': 'alex', 'exam 1': 'go', 'exam 2': 'C#'},
    {'name': 'steve', 'exam 1': 'C#', 'exam 2': 'go'}
]

I have tried the following:

for i, j in zip(a, b):
   if i['name'] == j['name']:
      c.update(i)

The result I am getting is just one dictionary that adds those lines that are the same not if their location is different.

CodePudding user response：

You could create two lookup dictionaries for each list and then find the intersection:

import pprint

lookup_A = {d["name"]: d for d in A}
lookup_B = {d["name"]: d for d in B}

result = [{**value, **lookup_B.get(key, {})} for key, value in lookup_A.items()]
pprint.pprint(result)

Output

[{'exam 1': 'python', 'exam 2': 'django', 'name': 'john'},
 {'exam 1': 'java', 'exam 2': 'C#', 'name': 'chris'},
 {'exam 1': 'C', 'exam 2': 'java', 'name': 'newman'},
 {'exam 1': 'C#', 'exam 2': 'C', 'name': 'sebast'},
 {'exam 1': 'python', 'exam 2': 'python', 'name': 'penier'},
 {'exam 1': 'go', 'exam 2': 'C#', 'name': 'alex'},
 {'exam 1': 'C#', 'exam 2': 'go', 'name': 'steve'}]

CodePudding user response：

With pandas you can do merge

pd.DataFrame(A).merge(pd.DataFrame(B), on='name').to_dict('records')

[{'name': 'john', 'exam 1': 'python', 'exam 2': 'django'},
 {'name': 'chris', 'exam 1': 'java', 'exam 2': 'C#'},
 {'name': 'newman', 'exam 1': 'C', 'exam 2': 'java'},
 {'name': 'sebast', 'exam 1': 'C#', 'exam 2': 'C'},
 {'name': 'penier', 'exam 1': 'python', 'exam 2': 'python'},
 {'name': 'alex', 'exam 1': 'go', 'exam 2': 'C#'},
 {'name': 'steve', 'exam 1': 'C#', 'exam 2': 'go'}]

CodePudding user response：

Though inefficient, here's a simple approach that is easy to follow, I suppose something like this was what you had in mind with your attempt using zip.

for a in A:
    for b in B:
        if b['name'] == a['name']:
            a.update(b)
    C.append(a)

CodePudding user response：

I would iterate over the lists and do a lookup before adding as follows:
Benefit is greater control. Cost is O(n) where n is the length of b

# Start off with values being a
c_dict = {x['name']:x for x in a}

# 'Upsert' from new list of dicts
for x in b:
    # You can update this to make this skip items not found in b
    # or other similar operations
    if not c_dict.get(x['name']):
         c_dict[x['name']] = x
    else:
        c_dict[x['name']].update(x)

# Finally convert to a list
c = list(c_dict.values())