skip to navigation
skip to content

jsontableschema-pandas 0.2.0

Generate Pandas data frames, load and extract data, based on JSON Table Schema descriptors.

Generate and load Pandas data frames based on JSON Table Schema descriptors.

Version v0.2 contains breaking changes:
  • removed Storage(prefix=) argument (was a stub)
  • renamed Storage(tables=) to Storage(dataframes=)
  • renamed Storage.tables to Storage.buckets
  • changed to read into memory
  • added Storage.iter to yield row by row

Getting Started


$ pip install datapackage
$ pip install jsontableschema-pandas


You can easily load resources from a data package as Pandas data frames by simply using datapackage.push_datapackage function:

>>> import datapackage

>>> data_url = ''
>>> storage = datapackage.push_datapackage(data_url, 'pandas')

>>> storage.buckets

>>> type(storage['data___data'])
<class 'pandas.core.frame.DataFrame'>

>>> storage['data___data'].head()
             Name Code
0     Afghanistan   AF
1   Åland Islands   AX
2         Albania   AL
3         Algeria   DZ
4  American Samoa   AS

Also it is possible to pull your existing data frame into a data package:

>>> datapackage.pull_datapackage('/tmp/datapackage.json', 'country_list', 'pandas', tables={
...     'data': storage['data___data'],
... })


Package implements Tabular Storage interface.

We can get storage this way:

>>> from jsontableschema_pandas import Storage

>>> storage = Storage()

Storage works as a container for Pandas data frames. You can define new data frame inside storage using storage.create method:

>>> storage.create('data', {
...     'primaryKey': 'id',
...     'fields': [
...         {'name': 'id', 'type': 'integer'},
...         {'name': 'comment', 'type': 'string'},
...     ]
... })

>>> storage.buckets

>>> storage['data'].shape
(0, 0)

Use storage.write to populate data frame with data:

>>> storage.write('data', [(1, 'a'), (2, 'b')])

>>> storage['data']
id comment
1        a
2        b

Also you can use tabulator to populate data frame from external data file:

>>> import tabulator

>>> with tabulator.Stream('data/comments.csv', headers=1) as stream:
...     storage.write('data', stream)

>>> storage['data']
id comment
1        a
2        b
1     good

As you see, subsequent writes simply appends new data on top of existing ones.


Please read the contribution guideline:

How to Contribute


File Type Py Version Uploaded on Size
jsontableschema-pandas-0.2.0.tar.gz (md5) Source 2016-10-26 8KB
jsontableschema_pandas-0.2.0-py2.py3-none-any.whl (md5) Python Wheel py2.py3 2016-10-26 9KB