EarthAI Notebook is a geospatially enabled python notebook designed to enable data scientists to extract insights from pixels.
Building on top of the JupyterLab environment, EarthAI provides many additional benefits including:
Direct access to catalogs of data
EarthAI has direct API access to the EarthAI catalog. Through this API users can search for imagery, filter and sort based on the image metadata, and import the data in an analysis ready format. Users can invoke the search directly using the EarthAI API or they can copy a query from the Earth OnDemand User Interface.
Seamlessly scalable infrastructure
EarthAI allows users to select the infrastructure they need for conducting their analysis. Depending on your plan, you will have access to one or more launch profiles. EarthAI supports GPU nodes as well as larger servers, and Spark Clusters. This makes scaling up your analysis seamless.
Fully configured (yet customizable) geospatial analysis environment
One of the most frustrating aspects of conducting an analysis is managing all the code dependencies and interoperability of python libraries. EarthAI comes with all of the most frequently used geospatial libraries installed and tested so you don't have to manage this complexity. Simply import your library and write your code. Don't worry - if you have a less common library, you can install that too.
EarthAI Notebook was designed to work with Apache Spark (when needed). RasterFrames, Astraea's open-source core, manages the complexity of parallel computation on Spark while enabling analysts to interact with raster data in a DataFrame. With embedded access to the Spark UI, users can harness the full power of distributed computation, debug their code, and monitor performance.
Article is closed for comments.