Graphs-in-Rust

Learning features of rust by implementing graphs with increasing complexity and robustness

Graph v1

Graph is a struct with vec and vec Vertex is just an id: usize Edge is unweighted directed edge with to: usize and from: usize

Graph v2

Graph is a struct with HashSet<Vertex> and HashSet<Edge> vertexes are indexed with disallow duplicate vertexes and duplicate edges

Ideas

• make graph a trait. look at javax.util.Graph for ideas of methods
• make graph, vertex, and edge generic
• make graph, vertex, and edge a trait so different implentations can be used
• MatrixGraph and ListGraph are Graph trait implementations
• WeightedUndirectedEdge, WeightedDirectedEdge, UndirectedEdge, DirectedEdge are edge implementations
• vertex requires an identifier and maybe needs default values
• vertex and edge must be hashable for graphs with set implementations
• be able to get an immutable graph from a graph
• be able to print graphs effectively
• implement different graph types with Big O analysis for standard operations
• be able to add graphs together
  • implement add trait to work like union of nodes and union of edges
  • any vertex with id1==id2 is treated as same vertex
  • SCC constructor that takes a list of nodes and returns a graph with nodes full connected
  • can construct graphs efficiently by defining a graph as the union of other graphs
    • scc() returns a graph with edges filled in between every node
    • ring() returns a graph with edges between each consecutive node including an edge from last to first
    • let g: Graph = Graph::scc('a', 'b', 'c') + Graph::ring('a', 'd', 'e')
    • makes a graph with Edges = {(a,b), (b,c), (c,a), (a,d), (d,e), (e,a)}
    • what other types of graphs are there?
• how can you implement a DAG such that there are no invalid states
• https://github.com/nrc/r4cppp/blob/master/graphs/README.md