Cycle detection in graphs in Data Structures Theory - Step-by-Step Execution

Concept Flow - Cycle detection in graphs

Start at a node

↓

Mark node as visited

↓

Explore neighbors

↓

Is neighbor unvisited?

Yes→Recurse on neighbor

↓

Detect cycle if neighbor is in recursion stack

↓

Backtrack

↓

Repeat for all nodes

↓

Cycle found?

Yes→Report cycle

No↓

No cycle in graph

Start from each node, mark it visited, explore neighbors recursively, detect cycle if a neighbor is already in the current recursion path.

Execution Sample

Data Structures Theory

Graph: A-B-C-D with edge D->B
Start DFS at A
Visit A, mark visited
Visit B, mark visited
Visit C, mark visited
Visit D, mark visited
From D, neighbor B is in recursion stack -> cycle detected

This example shows DFS traversal detecting a cycle when revisiting a node in the current recursion path.

Analysis Table

Step	Operation	Current Node	Visited Set	Recursion Stack	Action	Cycle Detected
1	Start DFS	-	{}	{}	Start at node A	No
2	Visit node	A	{A}	{A}	Mark A visited and add to recursion stack	No
3	Visit neighbor	B	{A}	{A}	B unvisited, recurse	No
4	Visit node	B	{A,B}	{A,B}	Mark B visited and add to recursion stack	No
5	Visit neighbor	C	{A,B}	{A,B}	C unvisited, recurse	No
6	Visit node	C	{A,B,C}	{A,B,C}	Mark C visited and add to recursion stack	No
7	Visit neighbor	D	{A,B,C}	{A,B,C}	D unvisited, recurse	No
8	Visit node	D	{A,B,C,D}	{A,B,C,D}	Mark D visited and add to recursion stack	No
9	Visit neighbor	B	{A,B,C,D}	{A,B,C,D}	B already in recursion stack	Yes
10	Backtrack	D	{A,B,C,D}	{A,B,C}	Cycle detected, stop	Yes

💡 Cycle detected at step 9 because neighbor B is already in recursion stack

State Tracker

Variable	Start	After Step 2	After Step 4	After Step 6	After Step 8	After Step 10
Visited Set	{}	{A}	{A,B}	{A,B,C}	{A,B,C,D}	{A,B,C,D}
Recursion Stack	{}	{A}	{A,B}	{A,B,C}	{A,B,C,D}	{A,B,C}

Key Insights - 3 Insights

Why do we check if a neighbor is in the recursion stack instead of just the visited set?

Because a node in the recursion stack means it's part of the current path, so revisiting it indicates a cycle. The visited set alone includes all visited nodes, not just the current path (see execution_table step 9).

What happens when we backtrack from a node?

We remove the node from the recursion stack but keep it in the visited set to avoid revisiting it again (see execution_table step 10).

Can a cycle be detected if the graph is disconnected?

Yes, cycle detection runs DFS from every unvisited node to cover disconnected parts (concept_flow shows starting from each node).

Visual Quiz - 3 Questions

Look at the execution_table at step 6, what is the recursion stack?

A {A,B,C}

B {A,B}

C {A,B,C,D}

D {}

💡 Hint

Check the 'Recursion Stack' column at step 6 in the execution_table

At which step does the cycle get detected?

A Step 7

B Step 9

C Step 10

D Step 5

💡 Hint

Look for 'Cycle Detected' column marked 'Yes' in the execution_table

If node B was unvisited at step 9, what would happen next?

A Algorithm would stop immediately

B Cycle would still be detected

C DFS would recurse into B again

D Visited set would be cleared

💡 Hint

Refer to the 'Action' column at step 9 and the concept_flow about recursion on unvisited neighbors

Concept Snapshot

Cycle detection in graphs uses DFS.
Track visited nodes and recursion stack.
If a neighbor is in recursion stack, cycle exists.
Backtrack removes nodes from recursion stack.
Check all nodes for disconnected graphs.

Full Transcript

Cycle detection in graphs is done by starting a depth-first search (DFS) from each node. We keep track of nodes visited and nodes currently in the recursion stack (the current path). When exploring neighbors, if we find a neighbor already in the recursion stack, it means a cycle exists. We backtrack by removing nodes from the recursion stack but keep them in the visited set to avoid repeated work. This process repeats for all nodes to cover disconnected graphs. The example graph with nodes A, B, C, D and an edge from D to B shows cycle detection when revisiting B in the recursion stack during DFS from A.

Practice

1. What is the main purpose of cycle detection in a graph?

easy

A. To count the number of nodes

B. To find if there is a loop in the graph

C. To sort the nodes in ascending order

D. To find the shortest path between nodes

Solution

1. Step 1: Understand the concept of cycle detection

   Cycle detection checks if a graph contains any loops where you can start at a node and return to it by following edges.

2. Step 2: Identify the main goal

   The main goal is to find if such loops exist, which can cause problems like infinite loops in algorithms.

3. Final Answer:

   To find if there is a loop in the graph -> Option B

4. Quick Check:

   Cycle detection = find loops [OK]

Hint: Cycle detection means finding loops in graphs [OK]

Common Mistakes:

• Confusing cycle detection with sorting
• Thinking it counts nodes instead of finding loops
• Assuming it finds shortest paths

2. Which data structure is commonly used to detect cycles in a directed graph using DFS?

easy

A. Queue

B. Stack

C. Hash Set to track recursion stack

D. Priority Queue

Solution

1. Step 1: Recall DFS cycle detection method

   DFS explores nodes deeply and uses a recursion stack to track nodes currently in the path.

2. Step 2: Identify the data structure used

   A hash set or boolean array is used to track nodes in the recursion stack to detect back edges indicating cycles.

3. Final Answer:

   Hash Set to track recursion stack -> Option C

4. Quick Check:

   DFS cycle detection uses recursion stack tracking [OK]

Hint: Use a hash set to track nodes in current DFS path [OK]

Common Mistakes:

• Using queue instead of stack for DFS
• Not tracking recursion stack nodes
• Confusing with BFS cycle detection

3. Consider the directed graph edges: [(1, 2), (2, 3), (3, 4), (4, 2)]. Does this graph contain a cycle?

medium

A. Yes, there is a cycle involving nodes 2, 3, and 4

B. Yes, but only between nodes 1 and 2

C. No, it is acyclic

D. No, because node 1 has no incoming edges

Solution

1. Step 1: Trace the edges to find cycles

   Edges form path 1->2->3->4 and then 4->2, which loops back to node 2.

2. Step 2: Identify the cycle nodes

   The cycle is formed by nodes 2, 3, and 4 because you can go from 2 to 3 to 4 and back to 2.

3. Final Answer:

   Yes, there is a cycle involving nodes 2, 3, and 4 -> Option A

4. Quick Check:

   Edges 4->2 create cycle 2-3-4 [OK]

Hint: Look for edges that point back to earlier nodes [OK]

Common Mistakes:

• Ignoring the edge 4->2 that closes the cycle
• Thinking node 1's edges affect cycle
• Assuming no cycle if start node has no incoming edges

4. Given this DFS-based cycle detection pseudocode, what is the error?

function dfs(node):
  visited[node] = true
  for neighbor in graph[node]:
    if visited[neighbor]:
      return true
    if dfs(neighbor):
      return true
  return false

medium

A. It does not track nodes in the current recursion stack

B. It marks nodes as visited too late

C. It should use a queue instead of recursion

D. It returns false too early

Solution

1. Step 1: Analyze the visited marking

   The code marks nodes as visited but does not distinguish between nodes visited in current path and fully processed nodes.

2. Step 2: Identify missing recursion stack tracking

   Without tracking nodes in the current recursion stack, it cannot detect back edges properly, causing false negatives.

3. Final Answer:

   It does not track nodes in the current recursion stack -> Option A

4. Quick Check:

   Missing recursion stack tracking causes wrong cycle detection [OK]

Hint: Track recursion stack separately to detect cycles [OK]

Common Mistakes:

• Using only visited array without recursion stack
• Confusing visited with recursion stack
• Thinking recursion depth causes error

5. You have a task scheduling system represented as a directed graph where edges mean "task A must finish before task B starts." How can cycle detection help in this system?

hard

A. It counts the total number of tasks

B. It finds tasks that can run in parallel

C. It sorts tasks by their duration

D. It detects impossible schedules due to circular dependencies

Solution

1. Step 1: Understand task scheduling graph meaning

   Edges show dependencies; a cycle means tasks depend on each other in a loop.

2. Step 2: Identify the role of cycle detection

   If a cycle exists, the schedule is impossible because tasks wait on each other endlessly.

3. Final Answer:

   It detects impossible schedules due to circular dependencies -> Option D

4. Quick Check:

   Cycle detection finds circular dependencies [OK]

Hint: Cycles mean tasks depend on each other endlessly [OK]

Common Mistakes:

• Thinking cycle detection sorts tasks
• Assuming cycles allow parallel tasks
• Confusing cycle detection with counting tasks