datenguidepy 0.4.1

Provids easy access to German publically availible regional statistics

• Free software: MIT license

• Documentation: https://datenguidepy.readthedocs.io/

The package provides easy access to German publicly available regional statistics. It does so by providing a wrapper for the GraphQL API of the Datenguide project.

Features

Overview of available statistics and regions:

The package provides DataFrames with the available statistics and regions, which can be queried by the user without having to refer to expert knowledge on regional statistics or the documentation of the underlying GraphQL API

Build and Execute Queries:

The package provides the user an object oriented interface to build queries that fetch certain statistics and return the results as a pandas DataFrame for further analysis.

Automatic inclusion of relevant meta data

Queries automatically retrieve some meta data along with the actual data to give the user more convenient access to the statistics without having to worry about too many technichal details

Full fidelity data

The package provides full fidelity data access to the datenguide API. This allows all use cases to use precicely the data that they need if it is available. It also means that most data cleaning has to be done by the user.

Quick Start

Install

To use the package install the package (command line):

pip install datenguidepy

Minimal example

To see the package work and obtain a DataFrame containing some statistics, the followin constitutes a minimal example.

from datenguidepy import Query

q = Query.region('01')
q.add_field('BEV001')
result_df = q.results()

Complex examples

These examples is intendend to illustrate many of the package’s features at the same time. The idea is to give an impression of some of the possibilities. A more detailed explanation of the functionality can be found in the the rest of the documentation.

q = Query.region(['02','11'])
stat = q.add_field('BEVSTD')
stat.add_args({'year' : [2011,2012]})
stat2 = q.add_field('AI1601')
stat2.add_args({'year' : [2011,2012]})
q.results(
    verbose_statistics = True,
    add_units = True,
).iloc[:,:7]

	id	name	year	Verfügbares Einkommen je Einwohner (AI1601)	AI1601_unit	Bevölkerungsstand (BEVSTD)	BEVSTD_unit
0	02	Hamburg	2011	22695	EUR	1718187	Anzahl
1	02	Hamburg	2012	22971	EUR	1734272	Anzahl
0	11	Berlin	2011	18183	EUR	3326002	Anzahl
1	11	Berlin	2012	18380	EUR	3375222	Anzahl

q = Query.region('11')
stat = q.add_field('BEVSTD')
stat.add_args({
    'GES' : 'GESW',
    'statistics' : 'R12411',
    'NAT' : 'ALL',
    'year' : [1995,1996]
})
stat.add_field('GES')
stat.add_field('NAT')
q.results(verbose_enums = True).iloc[:,:6]

	id	name	GES	NAT	year	BEVSTD
0	11	Berlin	weiblich	Ausländer(innen)	1995	191378
1	11	Berlin	weiblich	Deutsche	1995	1605762
2	11	Berlin	weiblich	Gesamt	1995	1797140
3	11	Berlin	weiblich	Deutsche	1996	1590407
4	11	Berlin	weiblich	Ausländer(innen)	1996	195301
5	11	Berlin	weiblich	Gesamt	1996	1785708

Get information on fields and meta data

Get information on region ids

# from datenguidepy import get_regions

 get_regions()

Use pandas query() functionality to get specific regions. E.g., if you want to get all IDs on “Bundeländer” use. For more information on “nuts” levels see Wikipedia.

get_regions().query("level == 'nuts1'")

Get information on statistic shortnames

#  from datenguidepy import get_statistics

  get_statistics()
  # return statistical descriptions in English
  get_statistics(target_language = 'en')

Get information on single fields

You can further information about description, possible arguments, fields and enum values on a field you added to a query.

q = Query.region("01")
stat = q.add_field("BEV001")
stat.get_info()

Further information

For detailed examples see the notebooks within the use_case folder.

For a detailed documentation of all statistics and fields see the Datenguide API.

Credits

All this builds on the great work of Datenguide and their GraphQL API datenguide/datenguide-api

The data is retrieved via the Datenguide API from the “Statistische Ämter des Bundes und der Länder”. Data being used via this package has to be credited according to the “Datenlizenz Deutschland – Namensnennung – Version 2.0”.

This package was created with Cookiecutter and the audreyr/cookiecutter-pypackage project template.

History

0.1.0 (2019-10-07)

• First release on PyPI.

0.1.1 (2019-10-09)

• Cleanup of the first release regarding naming, authors and docs.

0.2.0 (2020-11-30)

• Added functionality to use meta data for displaying descriptive statistics names and enum values

0.2.1 (2020-05-17)

• Added functionality to display the units of a statistic along with the numerical value.

• Internally split the meta data extraction into technical meta data and meta data about the statistics. Implemented new defaults for the statistics meta data in order to account for changes in the datenguide API.

0.2.2 (2020-05-24)

• Fixed a critical bug in the package data perventing the pypi version to essentially stop working completely.

• Fixed a bug related to incorrectly displayed version number of the package.

0.3.0 (2020-06-24)

• renamed get_all_regions to get_regions in accordance with get_statistics

• changed the index column name of the DataFrame returnd by all_regions from id to region_id

• made the statstics column name the index in the DataFrame returned by get_statistics and renamed it to statistic

• added functionality to obtain a stored auto-translated version of the get_statistics descriptions (default is German, now machine translation is available in English)

• introduced a new helper function get_availability_summary containing a (pre-calculated) summary of available data for region_id, statistic pairs down to nut3 level.

0.3.1 (2020-07-14)

• Introduced a better error messages for queries that are run without a statistic

• Bug fixes related to enums and auto join functionality

0.4.0 (2021-01-23)

• Introduced better error messages in case of invalid regions

• Introduced duplicate removal as an option for standard query results * New default is to remove duplicates but can be turned of with an argument * Auto-joining of multiple statistics should work better now as duplicates are removed before the joining. * Purpouse is only to remove duplicates that that may exist for technichal API reasons. The Purpouse is not to filter the data for content. * Rows are only counted as duplaces if everything, including the data source is identical

0.4.1 (2021-08-01)

• Bugfixe