Graph representations (adjacency matrix vs list) in Data Structures Theory - Performance Comparison

Time Complexity: Graph representations (adjacency matrix vs list)

O(n) for adjacency matrix, O(degree(i)) for adjacency list

Understanding Time Complexity

When working with graphs, how we store connections affects how fast we can check or find neighbors.

We want to know how the time to do these tasks grows as the graph gets bigger.

Scenario Under Consideration

Analyze the time complexity of checking all neighbors of a node using two common graph storage methods.

// Adjacency Matrix
for (int j = 0; j < n; j++) {
  if (matrix[i][j] == 1) {
    // process neighbor j
  }
}

// Adjacency List
for (int neighbor : list[i]) {
  // process neighbor
}
    

This code checks all neighbors of node i using either a matrix or a list.

Identify Repeating Operations

Look at what repeats when finding neighbors of one node.

• Primary operation: Looping through possible neighbors.
• How many times: For matrix, loops n times; for list, loops degree(i) times (number of actual neighbors).

How Execution Grows With Input

As the graph grows, the matrix always checks all nodes, but the list only checks actual neighbors.

Input Size (n)	Adjacency Matrix Ops	Adjacency List Ops
10	10 checks	degree(i) checks (e.g., 3)
100	100 checks	degree(i) checks (e.g., 5)
1000	1000 checks	degree(i) checks (e.g., 10)

Pattern observation: Matrix cost grows with total nodes, list cost grows with actual neighbors.

Final Time Complexity

Time Complexity: O(n) for adjacency matrix, O(degree(i)) for adjacency list

This means matrix always checks all nodes, while list only checks real neighbors, making it faster for sparse graphs.

Common Mistake

[X] Wrong: "Adjacency matrix is always faster because it uses a simple 2D array."

[OK] Correct: Matrix checks every node even if no connection exists, so it can be slower when many nodes have few neighbors.

Interview Connect

Understanding these differences helps you choose the right graph storage for your problem and explain your choice clearly.

Self-Check

"What if the graph is dense, meaning most nodes connect to many others? How would the time complexity for adjacency list compare to adjacency matrix?"

Practice

1. Which graph representation uses a 2D grid to show connections between nodes?

easy

A. Incidence matrix

B. Adjacency matrix

C. Edge list

D. Adjacency list

Solution

1. Step 1: Understand adjacency matrix structure

   An adjacency matrix is a 2D grid where rows and columns represent nodes, and cells show if an edge exists.

2. Step 2: Compare with other representations

   Adjacency lists store neighbors in lists, not grids. Edge lists and incidence matrices differ in format.

3. Final Answer:

   Adjacency matrix -> Option B

4. Quick Check:

   2D grid = adjacency matrix [OK]

Hint: Matrix means grid; list means neighbors [OK]

Common Mistakes:

• Confusing adjacency list with matrix
• Thinking edge list is a grid
• Mixing incidence matrix with adjacency matrix

2. Which of the following is the correct way to represent an adjacency list in Python?

easy

A. graph = [[1, 2], 0, [0, 1]]

B. graph = [[0,1,0],[1,0,1],[0,1,0]]

C. graph = [(0,1), (1,2), (2,0)]

D. graph = {0: [1, 2], 1: [0], 2: [0, 1]}

Solution

1. Step 1: Identify adjacency list format

   An adjacency list maps each node to a list of its neighbors, often using a dictionary in Python.

2. Step 2: Check each option

   graph = {0: [1, 2], 1: [0], 2: [0, 1]} uses a dictionary with keys as nodes and values as neighbor lists, which is correct. graph = [[0,1,0],[1,0,1],[0,1,0]] is a matrix, C is an edge list, D incorrectly uses an integer 0 for node 1 instead of a list.

3. Final Answer:

   graph = {0: [1, 2], 1: [0], 2: [0, 1]} -> Option D

4. Quick Check:

   Dict with neighbors = adjacency list [OK]

Hint: Adjacency list uses dict with node keys [OK]

Common Mistakes:

• Choosing matrix format as list
• Confusing edge list with adjacency list
• Using integer instead of list for neighbors

3. Given the adjacency matrix below, which nodes are connected to node 1?
graph = [[0, 1, 0], [1, 0, 1], [0, 1, 0]]

medium

A. Nodes 0 and 2

B. Nodes 1 and 2

C. Nodes 0 and 1

D. Nodes 2 only

Solution

1. Step 1: Locate row for node 1

   Row 1 in the matrix is [1, 0, 1], representing edges from node 1 to nodes 0, 1, and 2.

2. Step 2: Identify connected nodes

   Values 1 indicate connection. Here, positions 0 and 2 have 1, so node 1 connects to nodes 0 and 2.

3. Final Answer:

   Nodes 0 and 2 -> Option A

4. Quick Check:

   Row 1 has 1s at 0 and 2 [OK]

Hint: Check row for node, 1 means connected [OK]

Common Mistakes:

• Confusing row and column indices
• Including node itself as connected
• Misreading zeros as edges

4. What is wrong with this adjacency list representation?
graph = {0: [1, 2], 1: [0, 3], 2: [0], 3: 1}

medium

A. Node 3's neighbors should be in a list

B. Node 1 has an invalid neighbor

C. Node 0 should not have neighbors

D. The graph should be an adjacency matrix

Solution

1. Step 1: Check format of neighbors for each node

   Nodes 0, 1, and 2 have neighbor lists. Node 3 has a single integer instead of a list.

2. Step 2: Identify correct adjacency list format

   Neighbors must always be in a list, even if only one neighbor exists, to keep consistent structure.

3. Final Answer:

   Node 3's neighbors should be in a list -> Option A

4. Quick Check:

   Neighbors must be lists [OK]

Hint: Neighbors always in lists, never single values [OK]

Common Mistakes:

• Ignoring single neighbor format
• Thinking adjacency list must be matrix
• Assuming neighbors can be integers

5. For a graph with 1000 nodes and only 10,000 edges, which representation is more memory efficient and why?

hard

A. Adjacency matrix, because it allows faster edge checks

B. Adjacency matrix, because it uses fixed size memory

C. Adjacency list, because it stores only existing edges

D. Adjacency list, because it stores all possible edges

Solution

1. Step 1: Calculate memory use for adjacency matrix

   An adjacency matrix for 1000 nodes uses 1000x1000 = 1,000,000 cells, regardless of edges.

2. Step 2: Calculate memory use for adjacency list

   An adjacency list stores only the 10,000 edges, so memory use is proportional to edges, much less than matrix.

3. Step 3: Compare efficiency

   Since edges are sparse compared to possible connections, adjacency list is more memory efficient.

4. Final Answer:

   Adjacency list, because it stores only existing edges -> Option C

5. Quick Check:

   Sparse graph = adjacency list efficient [OK]

Hint: Sparse graph? Use adjacency list for less memory [OK]

Common Mistakes:

• Choosing matrix for sparse graphs
• Confusing speed with memory use
• Thinking adjacency list stores all edges