array parsing (flatten, unnest), pivot, filtering (keep), rename
This template flow shows you how to work with OpenStreetMap data in public Bigquery datasets. It also illustrates several important Google Cloud Dataprep features such as custom SQL, flow and dataset parameterization.
This flow comes with detailed annotation of each step in the recipe. To use this template, follow the following steps:
1) Go to openstreetmap.org and type in a location you want to use as the center of your search. Make note of the latitude and longitude.
2) Follow the annotations in the flow and right click on the Openstreetmap data source and select “Edit Custom SQL” to see the parameters defined, and override the default latitude and longitude to what you have selected from step 1.
3) Choose the OSM parsing recipe and select “+” Create Output to run. Click “Run” to bring up additional configuration options. You can override all of the parameters that were defined in the data source in step 2. There’s also a new parameter called amenity that allows you to select the type of POIs (point of interest) such as restaurant, cafe, atm… etc. that you want to see in your output. You can also configure the distance to search in meters via the distance_meters parameter.
Fore more information, please check out this video demonstrating the use of this flow template in Google Cloud Dataprep.