CODEWITHSUNDEEP
pythonjsonpandasdictionarydataframe
AntsaR
AntsaR

Reputation: 390

Parsing json values in pandas read_json

I am reading a json file to get data. The json begins like this: 

{
"rows": [
    [
        2013,
        "Mazda",
        "3",
        "917723",
        7795,
        20895
    ],
    [
        2016,
        "Cadillac",
        "ATS",
        "DD16135D",
        54890,
        66000
    ]

I am trying to get the values in a row into separate columns. I have tried doing the following:

df = pd.read_json(path, orient='values')

and got the result:

    rows
0  [2013, Mazda, 3, 917723, 77...
1  [2016, Cadillac, ATS, DD161...
2  [2015, Mitsubishi, Outlander,...
3  [2016, BMW, 528I, 918058, 3...
4  [2015, Toyota, Venza, 91806...

Ultimately, I would like to have 7 columns in my dataframe, each column representing each item in a row. How can I achieve this?

Upvotes: 1

Views: 1615

Answers (1)

jpp
jpp

Reputation: 164813

You can read the json file into a dictionary and select the 'rows' key:

# replace data with open('data.json')
with data as f:
    df = pd.DataFrame(json.load(f)['rows'])

print(df)

      0         1    2         3      4      5
0  2013     Mazda    3    917723   7795  20895
1  2016  Cadillac  ATS  DD16135D  54890  66000

Setup

from io import StringIO
import json

data = StringIO("""{
"rows": [
    [
        2013,
        "Mazda",
        "3",
        "917723",
        7795,
        20895
    ],
    [
        2016,
        "Cadillac",
        "ATS",
        "DD16135D",
        54890,
        66000
    ]]}""")

Upvotes: 2

Related Questions