Algorithms / Condensation Dag
Least You Need to Know: Condensation DAGs and Component-Level Ordering
If each SCC is contracted to one node, the resulting graph is a **DAG**. This condensation DAG captures the one-way flow between strongly connected regions and is often the right object for topological reasoning after SCC decomposition.
The least you need to know
- Each SCC becomes one meta-node in the condensation graph.
- Edges between SCCs represent original directed edges crossing components.
- The condensation graph is always acyclic.
- Topological ordering applies cleanly after SCC contraction.
- Many dependency arguments are easier at the component level than at the raw vertex level.
Key notation
cond(G)
condensation graph of directed graph G
meta-node
one strongly connected component viewed as a single node
DAG
directed acyclic graph
Tiny worked example
- Suppose SCC `A` has an edge to SCC `B`, and `B` has an edge to SCC `C`.
- In the condensation graph, this becomes `A → B → C`.
- If a cycle among meta-nodes existed, then all those SCCs would actually be mutually reachable and should merge into one larger SCC.
- That contradiction is why the condensation graph must be acyclic.
Common mistakes
- Students sometimes draw duplicate parallel edges; the presence of an edge matters more than multiplicity for most reasoning.
- A cycle inside an SCC does not make the condensation graph cyclic.
- Meta-nodes correspond to SCCs, not arbitrary DFS trees.
How to recognize this kind of problem
- You want to reason about SCC dependencies or topological order after compression.
- The problem asks what happens after contracting strongly connected regions.
- Cycles should disappear once mutually reachable blocks are merged.