CODEWITHSUNDEEP
pythonjsonpython-3.xpandasdataframe
nsky80
nsky80

Reputation: 115

How to convert JSON data into specified Pandas DataFrame

I have a json data which looks like this:

"rows": [
        ["2019-08-02", 364, 209, 2, 2],
    ["2019-08-03", 386, 250, 2, 5],
    ["2019-08-04", 382, 221, 3, 1],
    ["2019-08-05", 361, 218, 1, 0],
    ["2019-08-06", 338, 205, 4, 0],
    ["2019-08-07", 353, 208, 2, 2],
    ["2019-08-08", 405, 223, 2, 2],
    ["2019-08-09", 405, 266, 2, 2],
    ["2019-08-10", 494, 288, 0, 1],
        ]

I wanted to be headers of data as(not included in JSON file) as

["day", "estimatedPeopleVisited", "bought", "gives_pfeedback", "gives_nfeedback"]

I tried following code for reading file:

f = pd.read_json("data1308.json")
print(f)

and this gives output like:

                    rows
0   [2019-08-02, 364, 209, 2, 2]
1   [2019-08-03, 386, 250, 2, 5]
2   [2019-08-04, 382, 221, 3, 1]
3   [2019-08-05, 361, 218, 1, 0]
4   [2019-08-06, 338, 205, 4, 0]
5   [2019-08-07, 353, 208, 2, 2]
6   [2019-08-08, 405, 223, 2, 2]
7   [2019-08-09, 405, 266, 2, 2]
8   [2019-08-10, 494, 288, 0, 1]

I expect the output in form of:

       day      est   bought   gives_pfeedback    gives_nfeedback
0  2019-08-02   364    209           2                   2
1  2019-08-03   386    250           2                   5
2  2019-08-04   382    221           3                   1
3  2019-08-05   361    218           1                   0
4  2019-08-06   338    205           4                   0
.        .       .      .            .                   .
.        .       .      .            .                   .
.        .       .      .            .                   .

I can transform data in specified form after reading as problemset format but, is there any way to read directly JSON data in specified format?

Upvotes: 0

Views: 55

Answers (1)

ivallesp
ivallesp

Reputation: 2202

What about this?

import pandas as pd

data = {"rows": [
                 ["2019-08-02", 364, 209, 2, 2],
                ["2019-08-03", 386, 250, 2, 5],
                ["2019-08-04", 382, 221, 3, 1],
                ["2019-08-05", 361, 218, 1, 0],
                ["2019-08-06", 338, 205, 4, 0],
                ["2019-08-07", 353, 208, 2, 2],
                ["2019-08-08", 405, 223, 2, 2],
                ["2019-08-09", 405, 266, 2, 2],
                ["2019-08-10", 494, 288, 0, 1],
                    ]}
cols = ["day", "estimatedPeopleVisited", "bought", "gives_pfeedback", "gives_nfeedback"]

df = pd.DataFrame.from_dict(data["rows"])  
df.columns = cols

Upvotes: 3

Related Questions