RecallWeave / graph labA complete offline page

Algorithms · shortest paths

One decision closer
to a shortest path.

Discover a route, check whether it can improve, and see why settling the minimum makes its distance final. Every step uses the directed edges and weights you can see.

Original example.

Follow the arrows

0 settled
Unreached Tentative Settled Current target route

Settle, then relax

Distance from the source

VertexCostViaState

∞ means no route has been found. It becomes “unreachable” only when the discovered frontier is empty. “Via” is the recorded predecessor.

Change an edge. Start a new run.

Whole-number weights from 0 to 50. Zero is allowed. Any edit clears the old decisions; a negative or incomplete weight pauses the explorer.

What this step changed

    Earlier decisions

      Keep the run. Take the course.

      The course has 12 original questions across 4 connected concepts. Its fixed examples stay the same when you edit this graph. Open its JSON in RecallWeave’s Deck Studio to inspect or edit it; the separate lesson importer is still being qualified.

      A run download records this graph, every completed step and the current route status. It is a readable JSON record, not a saved RecallWeave learning session.

      The guarantee has a condition. Dijkstra’s settlement rule requires nonnegative edge weights. A target can be reachable while its distance is still tentative. Equal-cost routes may both be shortest; this trace breaks equal-distance vertex ties alphabetically and keeps the first predecessor on equal route cost.