Transcript Document
Tom Rink Tom Whittaker Paolo Antonelli Kevin Baggett Overview Local Hydra NetCDF, HDF Remot e DODS NetCDF, HDF VisAD - Jython VisAD Java Java3D VisAD ● Application Programming Interface (API) written in pure Java. ● Support for remote collaboration. ● Abstractions for: ● Data ● Displays ● Computations ● User Interface VisAD Data ● ● General mathematical representation of data; can model most any numerical data set. Can adapt data to this model from wide range of file storage formats: NetCDF, HDF-5, FITS, McIDAS and Vis5D. ● Finite samples of continuous functions. ● Supports mathematical operations and resampling of data. ● Built-in Metadata: ● Data schema (MathType) ● Physical Units ● Coordinate System transformations ● Sampling topology ● Error Estimates ● ● 2D or 3D interactive displays which support user direct manipulation: user can manually change display objects with changes propagated back to the linked data object. VisAD data and display class hierarchy includes both local and remote implentations so local and remote data look the same to applications. This is accomplished using Java's RMI (Remote Method Invocation) interface in VisAD core. Jython A pure Java implentation of Python http://www.jython.org ● Scripting environment (less code) ● Like Python, has easy to read syntax. ● ● ● Seamless access to Java classes (easy use of existing Java libraries in Jython modules and classes). No compile phase: Dynamic typing/loading. Generally leads to less typing and decreased overall development time. This leads into VisAD – Jython “Marriage” between the VisAD API and Jython ● ● Collection of Jython modules for the simplified creation of VisAD data and interactive displays. Jython is object-oriented so more complex applications, like Hydra, can also be developed in this environment. Hydra's To Do list: ● ● ● Add support for Geostationary instruments, eg. GOES, MSG, including animation over Time. Integrate multiple instruments into a single display, eg. overlay satellite swath with a flight track and toggle/difference between them. Remote collaboration: work together on a dataset over the network. ● Add more tools, image statistics etc. ● Performance tuning.