FSQ Spatial H3 Hub

The FSQ Spatial H3 Hub transforms geospatial data into analysis-ready tabular datasets using the H3 hexagonal grid system, eliminating traditional barriers for data scientists. Built specifically for data scientists, this innovative solution transforms geospatial data from complex raster pixels and vector polygons into analysis-ready tabular datasets using the standardized H3 hexagonal grid system. By pre-indexing data to H3 cells, we've eliminated the need for specialized geospatial tools or expertise, making it seamless to enrich your datasets with powerful spatial features. At launch, our Hub offers 20+ open datasets at resolution 8 delivered as an Iceberg Catalog, accessible directly from your preferred compute framework—whether you're working in Spark, Python, or DuckDB.

H3 Attributes

The following are some attributes that you will find in the H3 indexed datasets, that are generated as an artifact of our indexing process, along with the mapped attributes from the source datasets.

Attribute Name	Description
cellId	H3 cell id in integer format for specified resolution
cellCoverage	The proportion of the H3 cell area that intersects with the original polygon in vector format.
centroidCell	Boolean indicating whether or not the H3 cell contains the centroid of the polygon in vector format.
overlapGeom	Representation of the intersection between the original polygon and the H3 cell provided as WKB, in WGS84 (EPSG:4326). Filled only for rows where a cell is partially intersecting a polygon, meaning the cell is not fully contained by the polygon or the polygon is not fully contained by the cell
polygonCoverage	The proportion of the original polygon area (in vector format) that intersects with the H3 cell
pixelCoverage	The proportion of the raster pixel area that intersects with the H3 cell

Accessing the FSQ Spatial H3 Hub

The FSQ Spatial H3 Hub is your gateway to a variety of spatial datasets.

Navigate to the Hub: Open your web browser and go to the FSQ Spatial H3 Hub URL.
Browse Datasets: The main page displays a collection of available datasets. You can search for the datasets by name or filter them by various criteria such as their category, their source format, the geographical coverage or their owner.
Dataset Details: Once you find a dataset of interest, you can delve into the specifics of the dataset on its detail page. There you can examine the schema of the dataset and find the instructions to access the dataset from various compute engines.

Accessing the Iceberg Catalog

The datasets in the FSQ Spatial H3 Hub are organized in an Iceberg Catalog. To programmatically access data from the H3 Hub, you'll need an access token.

Navigate to Access Token Management:
1. From any dataset's "Sample Code" tab, locate and click on the "manage tokens" link or "Login to obtain your H3-Hub Token" under the "Prerequisites" section. This will typically lead you to a page where you can manage your access tokens.
2. Alternatively, you might find this under your user profile settings (e.g., "My Account" -> "Manage Tokens" or “My Account” -> Generate a new access token).
Generate a New Access Token:
1. Click on "generate a new access token".
2. Provide a descriptive Name for your token (e.g., "H3 analysis") and an optional description.
3. Select an Expiration period for the token (e.g., "1 day," "1 week," "1 month,"). Choose an appropriate duration based on your security practices.
4. Click "Generate Token". Your newly generated token will appear in a dialog box. Copy this token immediately as it will only be shown once.
Manage Existing Tokens:
1. On the "Manage Access Tokens" page, you can see a list of your existing tokens, their descriptions, expiration dates, and actions (e.g., delete). You can delete tokens that are no longer needed for security purposes.

Programmatic Access via Compute Engines

The "Sample Code" tab on the detail page of every dataset is a crucial resource for developers and data scientists. This tab provides ready-to-use code snippets and instructions tailored for accessing the selected dataset from various popular compute engines. You'll find guidance for:

DuckDB: Ideal for in-process SQL OLAP operations, allowing you to quickly query and analyze data directly within your application or local environment.
PySpark: For large-scale data processing and analytics, PySpark enables you to leverage Apache Spark's distributed computing capabilities to handle massive H3 indexed datasets.
PyIceberg: As the official Python client for Apache Iceberg, PyIceberg provides a robust way to interact with the Iceberg Catalog, allowing for efficient data access and management within your Python workflows.

Each section within the "Sample Code" tab typically includes prerequisites, instructions on how to configure your environment, and example queries to get you started with data retrieval and basic analysis.

Quick access from FSQ Spatial Workbench & FSQ Spatial Desktop

All datasets are automatically available in the MotherDuck instance that comes with your Workbench account—no additional setup required. You can also explore the datasets directly within Spatial Desktop by opening the pre-configured H3 Hub Example project.