Accumulate JSON Values in Python Dictionary as Array

user_3480969 0
Viewed 176

I have JSON File that has this format

{ 
"links": [
{"source":"0","target":"1","weight":1,"color":"white"},
{"source":"0","target":"2","weight":1,"color":"yellow"},
{"source":"0","target":"3","weight":1,"color":"white"},
]
}

I want to collect all target for a single source like this:

{"source": 0, "neighbors": ["1","2","3"]} where neighbors are all the collected target

Here's my Code

import json

with open("linksGr.json") as file:
    data = json.load(file)

collectDict = {}
for obj in data["links"]:
    if (collectDict["source"] == obj["source"]):
        collectDict["neighbour"] = obj["target"]

I just need a way to accumulate all targets for each source instead of there being multiple sources as I have done here 

collectDict["source"] = obj["source"]
collectDict["neighbour"] = obj["target"]

Any help would be appreciated a lot. I am sure there is some basic concept and a simple way that I am missing here. Thanks for the help.

dictionary python json
3 Answers

If I understand you correctly, you can use collections.defaultdict, to map from source to a list of targets, like this:

(I added some data to have multiple sources)

from collections import defaultdict

data = { 
"links": [
{"source":"0","target":"1","weight":1,"color":"white"},
{"source":"0","target":"2","weight":1,"color":"yellow"},
{"source":"0","target":"3","weight":1,"color":"white"},
{"source":"5","target":"7","weight":1,"color":"white"},
{"source":"5","target":"8","weight":1,"color":"yellow"},
{"source":"6","target":"9","weight":1,"color":"white"},
]
}

collectDict = defaultdict(list)
for obj in data["links"]:
    collectDict[obj["source"]].append(obj["target"])

print(dict(collectDict))

Output:

{'0': ['1', '2', '3'], '5': ['7', '8'], '6': ['9']}

EDIT: Here's another method using itertools.groupby, assuming the links are ordered by sources (otherwise, just sort it before)

from itertools import groupby

collectDict = {k: [t["target"] for t in g] for k,g in groupby(data["links"], lambda obj: obj["source"])}

print(collectDict)
user_50523650 
data = { 
    "links": [
        {"source":"0","target":"1","weight":1,"color":"white"},
        {"source":"0","target":"2","weight":1,"color":"yellow"},
        {"source":"0","target":"3","weight":1,"color":"white"},
        {"source":"5","target":"7","weight":1,"color":"white"},
        {"source":"5","target":"8","weight":1,"color":"yellow"},
        {"source":"6","target":"9","weight":1,"color":"white"},
    ]
}

collected = []
for obj in data["links"]:
    source_matches = [item for item in collected if item["source"] == obj["source"]]
    if len(source_matches) == 0:
        source_match = {"source": obj["source"], "neighbour": [obj["target"]]}
        collected.append(source_match)
    elif len(source_matches) == 1:
        source_matches[0]["neighbour"].append(obj["target"])
    else:
        raise BaseException()

print(collected)  # [{'source': '0', 'neighbour': ['1', '2', '3']}, {'source': '5', 'neighbour': ['7', '8']}, {'source': '6', 'neighbour': ['9']}]

Not very elegant, but does the job.

If you don't really need the format {"source": 0, "neighbors": ["1","2","3"]} I recommend the above solution using defaultdict. If you need this format, you could also build it from the output of the defaultdict solution.

user_76656910

Here is a simple solution:

import pandas as pd

result = [{"source": source, "neighbors": df["target"].tolist()}
          for source, df in pd.DataFrame(data["links"]).groupby("source")]
user_52961060
Related