Interpolation and Display of Randomly Scattered Scientific Data
Senior Project: 1992-1993
Andy Kuo, Dirk Mewes, Patricia Papic and Yasantha Samarasekera
This project was a three phased project. The first phase involved investigating
methods of interpolating 2D and 3D randomly spaced data onto 2D and 3D uniform
grids and the design and development of a C library for using some of these
methods. An emphasis was placed on choosing interpolation methods that provided
mechanisms for estimating errors and providing qualitative information about
the accuracy of the output grid. The second phase of the project was to build
an application using the interpolation library that allowed users, through an X
Windows Graphical User Interface, to interactively select data files, select
and configure interpolation methods, and finally produce an output grid file,
using the standard Network Common Data Format (netCDF). This application also
displayed qualitative information about the results of the interpolation,
allowing users to choose an interpolation method that best approximates their
data. The third phase was to develop with NCAR Graphics a utility, functioning
as a UNIX command with command line options, capable of reading and displaying
the gridded data, the random data and the error in the approximation.
|