In an earlier post entitled "Musing about Mazes" I gave some thought to how represent a maze, and some ideas about how to go about solving it and optimizing the mouse's path. Last night, while I was busy cleaning up my home office to make room to work on Maxwell, I gave it some more thought. I wasn't really happy with some of the 3D visuals of the maze/path in my previous post.
Fortunately I built the original maze representations as collections of cells in an Excel spreadsheet. From a formating perspective it would have been better to use a pure graphics/drawing program like Canvas, but for analytical purposes Excel was the best solution at hand.
A little experimentation with Excel's graphing functionality resulted in a dramatic improvement in the visualization of the maze. I switched the original 3D graph format from a 3D bar over to a 3D surface. For that format, Excel tends to use color bands, so I changed the colors to give the visual impression of a smooth color transitiona that match the path distance of each cell from the center of the maze.
For me, not for the mouse, this graph is much, much easier to work with. Looking at it, I have no trouble imagining a small marble rolling down the slopes, finding it's way to the finish cells in the center of the maze.
Hopefully it will provide me with some significant insights when I get to the point of actually developing the search and optimization algolrithms.