Scribd
Upload
16 views

Handling JSON

Uploaded by

Pranay Reddy
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
16 views

Handling JSON

Uploaded by

Pranay Reddy
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

How to Handle JSON File

import pandas as pd

x = '{"Name":["p1","p2"],"Age":[21,23]}'
## default json format simplest json

pd.read_json(x)

C:\Users\phane\AppData\Local\Temp\ipykernel_21116\1267235832.py:1: FutureWarning: Passing literal json to 'read_


json' is deprecated and will be removed in a future version. To read from a literal string, wrap it in a 'String
IO' object.
pd.read_json(x)
Name Age

0 p1 21

1 p2 23

x = '{1:["p1","p2"],"Age":[21,23]}' ## keys only take string datatype


## Can't use int,float,complex etc only strings

pd.read_json(x)

C:\Users\phane\AppData\Local\Temp\ipykernel_21116\1267235832.py:1: FutureWarning: Passing literal json to 'read_


json' is deprecated and will be removed in a future version. To read from a literal string, wrap it in a 'String
IO' object.
pd.read_json(x)
---------------------------------------------------------------------------
ValueError Traceback (most recent call last)
Cell In[13], line 1
----> 1 pd.read_json(x)

File ~\AppData\Roaming\Python\Python311\site-packages\pandas\io\json\_json.py:804, in read_json(path_or_buf, ori


ent, typ, dtype, convert_axes, convert_dates, keep_default_dates, precise_float, date_unit, encoding, encoding_e
rrors, lines, chunksize, compression, nrows, storage_options, dtype_backend, engine)
802 return json_reader
803 else:
--> 804 return json_reader.read()

File ~\AppData\Roaming\Python\Python311\site-packages\pandas\io\json\_json.py:1014, in JsonReader.read(self)


1012 obj = self._get_object_parser(self._combine_lines(data_lines))
1013 else:
-> 1014 obj = self._get_object_parser(self.data)
1015 if self.dtype_backend is not lib.no_default:
1016 return obj.convert_dtypes(
1017 infer_objects=False, dtype_backend=self.dtype_backend
1018 )

File ~\AppData\Roaming\Python\Python311\site-packages\pandas\io\json\_json.py:1040, in JsonReader._get_object_pa


rser(self, json)
1038 obj = None
1039 if typ == "frame":
-> 1040 obj = FrameParser(json, **kwargs).parse()
1042 if typ == "series" or obj is None:
1043 if not isinstance(dtype, bool):

File ~\AppData\Roaming\Python\Python311\site-packages\pandas\io\json\_json.py:1173, in Parser.parse(self)


1172 def parse(self):
-> 1173 self._parse()
1175 if self.obj is None:
1176 return None

File ~\AppData\Roaming\Python\Python311\site-packages\pandas\io\json\_json.py:1366, in FrameParser._parse(self)


1362 orient = self.orient
1364 if orient == "columns":
1365 self.obj = DataFrame(
-> 1366 ujson_loads(json, precise_float=self.precise_float), dtype=None
1367 )
1368 elif orient == "split":
1369 decoded = {
1370 str(k): v
1371 for k, v in ujson_loads(json, precise_float=self.precise_float).items()
1372 }

ValueError: Key name of object must be 'string' when decoding 'object'

simplest form of json is


where keys -- acting as column names

where values --- acting as column values

x = '{"Name":["p1","p2"],"Age":[21,23]}'

Different formats of json

x = '{"Name":["p1","p2"],"Age":[21,23]}'

data=pd.read_json(x)

C:\Users\phane\AppData\Local\Temp\ipykernel_21116\1694729012.py:1: FutureWarning: Passing literal json to 'read_


json' is deprecated and will be removed in a future version. To read from a literal string, wrap it in a 'String
IO' object.
data=pd.read_json(x)

data

Name Age

0 p1 21

1 p2 23

x2=data.to_json(orient="index")

x2

'{"0":{"Name":"p1","Age":21},"1":{"Name":"p2","Age":23}}'

pd.read_json(x2,orient="index")

C:\Users\phane\AppData\Local\Temp\ipykernel_21116\1222398037.py:1: FutureWarning: Passing literal json to 'read_


json' is deprecated and will be removed in a future version. To read from a literal string, wrap it in a 'String
IO' object.
pd.read_json(x2,orient="index")
Name Age

0 p1 21

1 p2 23

x3 =data.to_json(orient="columns")

x3

'{"Name":{"0":"p1","1":"p2"},"Age":{"0":21,"1":23}}'

pd.read_json(x3,orient="columns")

C:\Users\phane\AppData\Local\Temp\ipykernel_21116\2742886167.py:1: FutureWarning: Passing literal json to 'read_


json' is deprecated and will be removed in a future version. To read from a literal string, wrap it in a 'String
IO' object.
pd.read_json(x3,orient="columns")
Name Age

0 p1 21

1 p2 23

x4 = data.to_json(orient="values")

x4

'[["p1",21],["p2",23]]'

pd.read_json(x4,orient="values")

C:\Users\phane\AppData\Local\Temp\ipykernel_21116\2664431613.py:1: FutureWarning: Passing literal json to 'read_


json' is deprecated and will be removed in a future version. To read from a literal string, wrap it in a 'String
IO' object.
pd.read_json(x4,orient="values")
0 1

0 p1 21

1 p2 23

x5=data.to_json(orient="split")

x5

'{"columns":["Name","Age"],"index":[0,1],"data":[["p1",21],["p2",23]]}'

pd.read_json(x5,orient="split")

C:\Users\phane\AppData\Local\Temp\ipykernel_21116\426387458.py:1: FutureWarning: Passing literal json to 'read_j


son' is deprecated and will be removed in a future version. To read from a literal string, wrap it in a 'StringI
O' object.
pd.read_json(x5,orient="split")
Name Age

0 p1 21

1 p2 23

Semi-Structured Json file


x = {"Name":"p1","marks":21} #semi_structure takes the format in dictionary
# whereas structured json takes the format in string

{'Name': 'p1', 'marks': 21}

type(x)

dict

x = {"Name":"p1","marks":{"Maths":21,"Hindi":20}}

pd.json_normalize(x)

Name marks.Maths marks.Hindi

0 p1 21 20

x = {"Name":"p1","marks":{"Sem1":{"Maths":21,"Hindi":20},"Sem2":{"Maths":22,"Hindi":25}}}

pd.json_normalize(x)

Name marks.Sem1.Maths marks.Sem1.Hindi marks.Sem2.Maths marks.Sem2.Hindi

0 p1 21 20 22 25

pd.json_normalize(x,max_level=0)

Name marks

0 p1 {'Sem1': {'Maths': 21, 'Hindi': 20}, 'Sem2': {...

pd.json_normalize(x,max_level=1)

Name marks.Sem1 marks.Sem2

0 p1 {'Maths': 21, 'Hindi': 20} {'Maths': 22, 'Hindi': 25}

pd.json_normalize(x,max_level=2)

Name marks.Sem1.Maths marks.Sem1.Hindi marks.Sem2.Maths marks.Sem2.Hindi

0 p1 21 20 22 25

x = {"Name":{"p1":12},"marks":[{"Sem1":20,"Sem2":30}],"Age":{"a":{"b":1}}}

pd.json_normalize(x,record_path=["marks"],meta=[["Name","p1"],["Age","a","b"]])
Sem1 Sem2 Name.p1 Age.a.b

0 20 30 12 1

Loading [MathJax]/jax/output/CommonHTML/fonts/TeX/fontdata.js

You might also like