Skip to content

Identity Matrix

Simple Idea

  • Think of the Identity Matrix as the number 1 in the world of matrix multiplication. Multiplying any matrix by the Identity matrix (of the right size) leaves the original matrix unchanged.

Formal Definition

  • The Identity Matrix, denoted as \(\mathbf{I}\) or \(\mathbf{I}_n\) (where \(n\) is the dimension), is a square matrix (\(n \times n\)) with 1s on the main diagonal (from top-left to bottom-right) and 0s everywhere else.

Key Concepts

1. Representation

  • Structure: \((\mathbf{I}_n)_{ij} = 1\) if \(i=j\) (on the diagonal), and \((\mathbf{I}_n)_{ij} = 0\) if \(i \neq j\) (off-diagonal). Using Kronecker delta: \((\mathbf{I}_n)_{ij} = \delta_{ij}\).
  • Examples: $$ \mathbf{I}_2 = \begin{bmatrix} 1 & 0 \ 0 & 1 \end{bmatrix} \quad \quad \mathbf{I}_3 = \begin{bmatrix} 1 & 0 & 0 \ 0 & 1 & 0 \ 0 & 0 & 1 \end{bmatrix} $$

2. Multiplicative Identity Property

  • For any \(m \times n\) matrix \(\mathbf{A}\): $$ \mathbf{I}_m \mathbf{A} = \mathbf{A} $$
  • For any \(n \times p\) matrix \(\mathbf{B}\): $$ \mathbf{B} \mathbf{I}_p = \mathbf{B} $$
  • Requirement: The Identity matrix must have compatible dimensions for the matrix multiplication to be defined. \(\mathbf{I}_n\) multiplies \(n \times p\) matrices on the left, \(\mathbf{I}_n\) multiplies \(m \times n\) matrices on the right.

3. Role in Linear Transformations

  • As a transformation, the Identity matrix represents "doing nothing". Applying \(\mathbf{I}\) to a vector \(\mathbf{x}\) results in the same vector \(\mathbf{x}\): \(\mathbf{Ix = x}\).

Connections to Other Topics

  • Crucial for defining the Matrix Inverse (\(\mathbf{A}^{-1}\)), where \(\mathbf{A} \mathbf{A}^{-1} = \mathbf{A}^{-1} \mathbf{A} = \mathbf{I}\).
  • Used in solving systems of linear equations \(\mathbf{Ax = b}\).
  • Appears in various matrix decompositions.

Summary

  • The Identity Matrix (\(\mathbf{I}\)) is the matrix equivalent of the number 1 for multiplication.
  • Square matrix with 1s on diagonal, 0s elsewhere.
  • \(\mathbf{IA = A}\) and \(\mathbf{AI = A}\) (when dimensions match).
  • Represents a "do nothing" linear transformation.

Sources