The dataMineLib provides data mining capabilities to Analytic Information Server applications. It is comprised of four primary components:
The dataMineLib object repository manager provides you with the ability to manage up to 100 repository extents. Each of these repository extents can contain up to 2 gigabytes or 8 million objects. The dataMineLib provides an interface for creating and manipulating table objects in each repository extent. Each table is comprised of rows of object vectors of variable size. The elements of each object vector may be accessed by "column" name. Each row column may be any kind of Analytic Information Server immediate or heap object. See the dataMineLib Interface section for an overview of the dataMineLib and a detailed reference for each dataMineLib object function and property.
The dataMineLib cursor Lambda provides you with the ability to open and manipulate tables through a table cursor object. Table cursor objects provide a rich interface for manipulating tables. See the Table Cursors Section for an overview of table cursors and a detailed reference on each cursor object function and property.
The javaScript data filtering language is a sophisticated query scripting language that you use to analyze information in a dataMineLib table. The javaScript language is very similar to JavaScript. The javaScript language has been enhanced with keywords and functions that give it direct access to dataMineLib tables. See the javaScript Lambda documentation for an overview and detailed reference on this powerful language.
The dataMineLib.miner data mining project manager uses the dataMineLib object repository to manage a collection of tables that contain a data mining project. These project tables can include both data and javaScript programs. See the Miner Section for an introduction and detailed reference for the miner Lambda.
This document describes the dataMineLib, dataMineLib table cursor object, dataMineLib.miner Lambdas in detail.