A visual programming language for visualization of scientific data
Daniel David Hils · 1992
Today scientists are confronted with the problem of understanding and analyzing large masses of numeric scientific data. One solution to this problem is scientific visualization: converting the numeric data into pictures more readily understood by scientists. This thesis presents DataVis, a visual programming language that is designed to be used by scientists for the visualization of scientific data. Since DataVis is a visual language, DataVis programs and functions are primarily diagrams, rather than text. The language is based on the data flow computational model. DataVis provides a library of predefined programs and functions. Scientists transform data into pictures by connecting data icons to program icons. The program icons can be customized, extended, and composed using the same plumbing metaphor, and new program icons can be defined using the same metaphor. Thus, DataVis is both a set of finished programs for visualizing scientific data and an easy-to-use toolset for constructing more. It is a complete programming language, so there is no limit to the kinds of visualization tools that it can construct, but its programming language is built on the same metaphor as the user interface of its applications, so it is easy for non-programmers to learn.