Deepak Cherian, Scott Henderson, & Emma Marshall - Tidy Geospatial Cubes | SciPy 2023

Join experts as they explore Tidy Geospatial Cubes, applying array data tidying principles to satellite imagery datasets with a focus on metadata and community-driven frameworks.

Key takeaways
  • The talk discusses the concept of “Tidy Geospatial Cubes” and its application to array data
  • The presentation uses examples of satellite imagery datasets to illustrate the importance of tidying data for analysis
  • The speakers introduce the concept of x-ray and its role in structuring data cubes
  • The talk highlights the challenges of working with large, complex geospatial datasets and the need for a standardized framework for tidying data
  • The speakers emphasize the importance of metadata and its role in describing the structure of a dataset
  • Tidy principles for array data could look like, including:
    • Each variable forms a column
    • Each observation forms a row
    • Each type of observational unit forms a table
  • The talk introduces the concept of CF conventions for ancillary data and coordinate types
  • The speakers discuss the need for a community-driven framework for tidying data and invite feedback and contributions from users
  • The presentation concludes by emphasizing the importance of collaboration and community knowledge in developing a standardized framework for tidying data