Skip to content
The missing datasets manager. Like hombrew but for datasets. CLI-tool for search and discover datasets!
Python Makefile
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
docs
src/databrewer
tests
.bumpversion.cfg
.coveragerc
.editorconfig
.gitignore
.travis.yml
AUTHORS.rst
CONTRIBUTING.rst
HISTORY.rst
LICENSE
MANIFEST.in
Makefile
README.rst
TODO.rst
VERSION
pytest.ini
requirements-dev.txt
requirements-install.txt
requirements-setup.txt
requirements-tests.txt
setup.cfg
setup.py
tox.ini

README.rst

DataBrewer

Documentation Status Coverage Status Code Quality Status Requirements Status

The missing datasets manager.

DataBrewer preview

Databrewer let you search and discover datasets. Inspired by Homebrew, it creates and index of known datasets that you can download with a single command. It will provide an API to allow to do the same in, for example, a IPython notebook so you no longer have to manually download datasets.

Quickstart

Install databrewer:

pip install databrewer

Update the recipes index:

databrewer update

Search for some keywords:

databrewer search nyc taxi

Example output:

andresmh-nyc-taxi-trips - NYC Taxi Trips. Data obtained through a FOIA request
nyc-tlc-taxi            - This dataset includes trip records from all trips
                          completed in yellow and green taxis in NYC in 2014 and
                                                    select months of 2015.

Let's check the nyc-tlc-taxi dataset:

databrewer info nyc-tlc-taxi

We can either download the entire dataset (which is huge!):

databrewer download nyc-tlc-taxi

Or just a few files in the dataset, or select a subset:

databrewer download "nyc-tlc-taxi[green][2014-*]"

Note

Note that * is the standard glob operator and [green] acts as selector. The selectors depends on how the recipe if defined. When using selectors you must enclose the name in quotes in most shells.

Finally you need to know where the files are located for further processing:

databrewer download "nyc-tlc-taxi[green][2014-*]"

Example output:

/Users/rolando/.databrewer/datasets/nyc-tlc-taxi/green_tripdata_2014-01.csv
/Users/rolando/.databrewer/datasets/nyc-tlc-taxi/green_tripdata_2014-02.csv
/Users/rolando/.databrewer/datasets/nyc-tlc-taxi/green_tripdata_2014-03.csv
/Users/rolando/.databrewer/datasets/nyc-tlc-taxi/green_tripdata_2014-04.csv
/Users/rolando/.databrewer/datasets/nyc-tlc-taxi/green_tripdata_2014-05.csv
/Users/rolando/.databrewer/datasets/nyc-tlc-taxi/green_tripdata_2014-06.csv
/Users/rolando/.databrewer/datasets/nyc-tlc-taxi/green_tripdata_2014-07.csv
/Users/rolando/.databrewer/datasets/nyc-tlc-taxi/green_tripdata_2014-08.csv
/Users/rolando/.databrewer/datasets/nyc-tlc-taxi/green_tripdata_2014-09.csv
/Users/rolando/.databrewer/datasets/nyc-tlc-taxi/green_tripdata_2014-10.csv
/Users/rolando/.databrewer/datasets/nyc-tlc-taxi/green_tripdata_2014-11.csv
/Users/rolando/.databrewer/datasets/nyc-tlc-taxi/green_tripdata_2014-12.csv

Datasets

The aim is to index known and not-so-known datasets. There is no plans to standarize the dataset format as we want to keep it as published by the authors.

Recipes

Datasets are defined in recipes which contains information about the dataset and where to find it.

These recipes are community maintained and hosted in the databrewer-recipes repository.

Roadmap

  • Include an API. For now it only provides a CLI-interface but in the near future it will include an API so you can search, download and load datasets directly in your Python code.

Contributing

You can help by the following means:

See CONTRIBUTING.rst for more information.

You can’t perform that action at this time.