Talk:Data Structures/Sets

I think the Sets chapter should be before the Graph chapter because the Graph interface could benefit from returning edge sets and node sets from its functions. What needs to be changed to accomplish this? --jyasskin 16:31, 27 Jan 2005 (UTC)


 * We should stick with the adjacency matrix and adjacency list representations for graphs and present them before sets. A disjoint set will be radically different and is more advanced of a topic (and requires knowledge of trees; further, some non-disjoint sets should be efficient search trees themselves). To iterate over the edges requires only an edge iterator accessor. I suppose here, though, the design of BFS and DFS should dictate what will really be useful for the graphs chapter. We should simplify some of these interfaces and move closer to the Java version than the C++ library. MShonle 05:05, 28 Jan 2005 (UTC)