Forum Discussion

vgautam64's avatar
vgautam64
New Contributor III
2 years ago
Solved

Reduce an array of objects into a single object with the sum of fields as the result

I have an array of the following form: I want to convert this array of objects into a single object that has the same fields along with Debit_Amount and Credit_Amount fields summed up. ...
  • koryknick's avatar
    koryknick
    2 years ago

    vgautam64 - I believe this solves your question in the simplest way I can think of.  Use a Mapper snap with the following configuration:

    It appears to me that you only need the field elements and values from the first group element: $group[0]

    The two jsonPath statements are simply pulling all values for the Debit and Credit Amounts into an array of numbers

    Finally, the Array.reduce() method is used to aggregate that list of numbers into a final total.

    Hope this helps!

     

Supratim's avatar
Supratim
Contributor III
2 years ago

vgautam64  As I can't copy your json, have created my own, you can use it- 

Data  ==> 

{
    "carList": [
    {
    "color": "purple",
    "type": "minivan",
    "debit_amount": 50,
    "credit_amount": 7
  },
  {
    "color": "red",
    "type": "station wagon",
    "debit_amount": 100,
    "credit_amount": 5
  }, {
    "color": "black",
    "type": "toyota",
    "debit_amount": 200,
    "credit_amount": 5
  },
  {
    "color": "blue",
    "type": "Maruti",
    "debit_amount": 300,
    "credit_amount": 5
  }
]
}
 
1. mapper  : 
$carList.map((val,index)=>{"sum_debit_amount": jsonPath($, "carList[*].debit_amount").reduce((cur,acc)=>cur+acc),"sum_credit_amount" : jsonPath($, "carList[*].credit_amount").reduce((cur,acc)=>cur+acc)}.extend(val)).shift()
  • vgautam64's avatar
    vgautam64
    New Contributor III
    2 years ago

    Hi Supratim,

    Thanks for the providing a solution. I tried to validate the expression with the data I have and it seems to work when I validate the pipe. Strangely though, I observe that when I execute (not validate) the pipeline, the mapper that holds the expression only seem to process 50 records and gets stuck processing the 51st doc. But when I validate the pipe I am able to see the preview of 148 documents.

    Any idea why this would happen?

  • jaybodra's avatar
    jaybodra
    New Contributor III
    7 years ago

    Hi Jack,

    the best work around to this use json.loads and json.dumps; json.loads(in_doc.get(‘str_json’)). use a mapper with JSON.stringify() to represent JSON as string and then you can use json.loads() in the script.

  • Henchway's avatar
    Henchway
    Contributor
    5 years ago

    Have you ever found a good solution to this besides converting it to a string prior?

    Generally, the proposed method by @jaybodra works, but it seems more like a short-term workaround.
    Additionally, I’m unable to retain the sorting of the dumps after re-loading it, even when telling it to do so specifically. So this only leaves the option of entering a string and passing out a string and then using JSON.parse(), then it will actually be an ordered json … 😮

    json_input = json.loads(inDoc.get('content'))
json_string = json.dumps(json_input, sort_keys=True)
json_output = json.loads(json_string, object_pairs_hook=OrderedDict)

    Best regards
    Thomas