SciSpark is a scalable scientific processing platform that makes interactive computation and exploration possible. This prototype project outlines the architecture of the scientific RDD (SRDD), and a library of linear algebra and geospatial operations. Its initial focus is expressing Grab em' Tag em' Graph em' algorithm as a map-reduce algorithm.
The full installation procedure for SciSpark can be found here.
The steps to getting started with a simple SciSpark example can be found here.
The steps to running the SciSpark test suite can be found here.
SciSpark documentation can be found at https://scispark.jpl.nasa.gov/api/
SciSpark frontend interface leverages notebook software and is built on Apache Zeppelin. Examples of notebooks leveraging the SciSpark API and furnished with visualizations of the data analyzed can be found here.
The initial application of SciSpark is to scale the Grab em' Tag em' Graph em' algorithm. More about GTG can be found here.
Contact us at scispark-team@jpl.nasa.gov