AQL & AFL Query Languages

SciDB supports two query languages for accessing and analyzing data programmatically: a SQL-like Array Query Language (AQL), and an Array Functional Language (AFL) so you can use whichever fits your comfort zone. Each natively supports the following capabilities:

• Data schema definition
• Large-scale math operations embedded directly in queries
• Array data management (filter, aggregate, join, etc.)
• Parallel statistical & linear algebra (covariance, regression, PCA, etc.)
• User-defined operators (Postgres-style)

AQL: Array Query Language

AQL is a declarative language, modeled after SQL, and extended for working with array data. Here is an example of what AQL looks like.

/* a sample AQL query */
SELECT avg ( S.v1 )
  FROM Simple_Array S 
 WHERE S.v3 = ‘Odd’ 
 GROUP BY S.I;

For documentation about AQL and other features of SciDB, visit the downloads page on the SciDB forum.

AFL: Array Functional Language

AFL provides a more powerful functional language interface. Here is an example of the exact same query shown above, but written in AFL:

/* a sample AFL query */
aggregate( filter ( Simple_Array, v3 = ‘Odd’),
           I, avg ( Simple_Array.v1 ) );

For documentation about AFL and other features of SciDB, visit the downloads page on the SciDB forum.