CS 350: Algorithms for Geographic Information Systems

Spring 2004

T, Th 1:00 - 2:25 in Searles 126/224

Prerequisites: CS231-Algorithms (or permission of instructor)

Instructor: Laura Toma
Office: 219 Searles Hall
Email: ltomaATbowdoin.edu (replace AT with @)

Office hours: TBD; Also, any time I am in the office (and in a good mood :)). Or, just send me an email to setup a (different) time.

Class Email: csci350ATbowdoin.edu (replace AT with @)

Class webpage: http://www.bowdoin.edu/~ltoma/teaching/cs350/spring04
Bowdoin class webpage: http://academic.bowdoin.edu/courses/s04/csci350/

Course topics:

• Geographic data models and representation (points, lines, polygons, raster, TIN, contours). Transformations between models. Interpolation.
• Delaunay triangulation. Storing, accesssing and traversing a triangulation. Triangulation overlay and point location.
• Line/terrain simplification/generalization.
• Spatial data structures: Interval tree, B-tree, quad-tree, oct-tree, kd-tree, range-tree, grid file, R-tree. Dynamization.
• Range searching. Rectangle-rectangle searching. Nearest-neighbor queries. Window queries. Stabbing queries. Segment intersection.
• Topographic analysis of terrains. Flow modeling.
• Very large datasets. Motivation. Scalability of algorithms to large datasets. Models of computation. The I/O-model.
• External memory algorithms. External memory sorting. External memory priority queue. External memory basic graph algorithm (BFS, DFS, SSSP).

Course material:

• The C programming Language. Kernighan and Ritchie. Second edition. (Required.)
• Algorithmic Foundations of geographic Information Systems.. Kreveld, Nievergelt, Roos, Widmayer Editors. LNCS tutorial. (Optional.)
• Papers, handouts, manuals etc.

Other good reference books include:

• Computational geometry. M. de Berg, M. van Kreveld, M. Overmars, O. Schwarzkoph.
• Computational geometry in C. J. O'Rourke.
• Programming pearls. J. Bentley.

Grading policy

• Programming projects
• Class presentations

Programming projects

The final goal is to implement a primitive Geographic Information System. The projects will be challenging and will solve real-world problems on real-world data.

The projects will be made available online as the semester progresses and will be due roughly every two weeks. If you need an extension, just talk to me. I am well aware that programming time can vary deeply from person to person and from day to day. Rather than submitting bad or incomplete code, I prefer that you take extra time, do a good job and be happy with what you turn in. You can have at most two extensions the entire semester without affecting your grade.

As you probably all know by now, a big and sometimes frustrating part of programming is debugging your code. This is something that you have to go through. I will be there to help occasionally, but do not count on it and do not take it for granted.