Recipes: processing instructions#

Recipes are instructions for ingesting and transforming data.

They describe where the input data comes from and how it needs to be processed at every step (e.g., data ingestion, harmonization, curation, etc.).

Recipes make openplaces scalable: the package provides an abstraction logic, and the recipes translate external data to that logic.

Contributed recipes expand the analytical capacities of all other users.

Identifiers#

Recipes are identified by their recipe_id, a concatenation of:

Administrative units: geography (optional)
an identifier of Entities: the units of analysis or Datasets: the attributes, and
also optionally: a filename.

Examples:

Recipe ID	Description
`admin-openplaces-2026_admin1`	Global identifiers of countries/territories (administrative units, level 1: `admin1`). This recipe is a table
`admin-gadm-4~1_admin1`	Global administrative geometries from the Global Administrative Database (GADM) This recipe contains instructions on how to access the data
`US_admin-census-2021_admin3`	United States: official county (`admin3`) boundaries from the US Census
`US_footprint-microsoft-v2`	US building footprints provided by Microsoft
`US-NC_parcel-nconemap-2023`	Parcel boundary data for North Carolina, United States, provided by NCOneMap

Location#

Recipes live in src/openplaces/recipes.

They use the same directory structure as internal data directories: grouped by administrative units, followed by entities or datasets.

Understanding recipes#

For a browsable list of the available recipes and what each dataset contains, see the recipe catalog (generated from the recipes’ description fields).

For an in-depth treatment of recipe arguments and functionality, see instructions on writing recipes in the section for contributors.