tag:blogger.com,1999:blog-8599014026720999293.post6474497982211639584..comments2016-06-23T05:47:57.863-07:00Comments on Joe on Software: Trying to use math to solve my problemsJoe Heyminghttps://plus.google.com/104875384702314339067noreply@blogger.comBlogger3125tag:blogger.com,1999:blog-8599014026720999293.post-37103160182600283732014-10-28T15:47:11.040-07:002014-10-28T15:47:11.040-07:00How would that look using SVG?How would that look using SVG?Joe Heyminghttps://www.blogger.com/profile/11310359026426664481noreply@blogger.comtag:blogger.com,1999:blog-8599014026720999293.post-16882757355256577112014-10-08T09:22:30.138-07:002014-10-08T09:22:30.138-07:00It looks like you're trying to find an approxi...It looks like you're trying to find an approximate solution to Laplace's equation as a boundary value problem; more specifically a solution to the Dirchelet problem. This might help you direct your search.<br />An iterative solution would be to start with the initial "list of points" and make a higher-resolution grid which includes those points collectively as the boundary function. All the remaining non-boundary values in the grid could be evaluated as the average of their neighbor values. Such an algorithm would iterate until successive passes yield sufficiently low changes.The resulting higher-resolution grid would lend itself to connection as a contour map that would have the characteristic roundness you seek.stevehttps://www.blogger.com/profile/08164472119404195304noreply@blogger.comtag:blogger.com,1999:blog-8599014026720999293.post-69678727860243117022014-10-08T00:12:10.948-07:002014-10-08T00:12:10.948-07:00Have you tried looking at the problem from another...Have you tried looking at the problem from another end? Try rendering the polygons white on a black background, and the use a Gaussian blur with a large radius (you can try this out in Photoshop/GIMP). Then you can use the resulting grayscale image as a source to heatmap.Michal Kottmanhttps://www.blogger.com/profile/05590245710117442912noreply@blogger.com