Obit: A Development Environment for Astronomical Algorithms

Obit is a development environment for astronomical algorithms. It implements multiple "native" data formats including FITS and AIPS. The shared data model with AIPS means almost complete interoperability, allowing algorithmic development in a production environment. The Obit package supports both prepackaged, compiled tasks and a python interface to the major class functionality to allow rapid prototyping using python scripts. Developed features are predominantly radio astronomical, but support for astronomical images and tables is generic and should allow a wider range of applications. The design is intended to be readily extensible. An example is given that explores the effect of "CLEAN bias" in the VLA Low-Frequency Sky Survey and compares its method of CLEAN to an automated CLEAN window selection scheme that should reduce the bias. This technique shows much promise for reducing this systematic effect.