Row Operations and Matrix Inverses
What you'll learn: You'll discover how the elementary row operations you perform during Gaussian elimination are actually equivalent to multiplying by special matrices called elementary matrices.
The Core Connection
When you perform row operations to find a matrix inverse, you're not just manipulating numbers randomly—you're actually multiplying by specific matrices behind the scenes. Each elementary row operation (swapping rows, multiplying a row by a scalar, or adding a multiple of one row to another) corresponds to multiplication by an elementary matrix.
What is an Elementary Matrix?
An elementary matrix is what you get when you perform a single elementary row operation on the identity matrix. For example:
- Row swap: Swapping rows 1 and 2 of the identity gives you a swap matrix
- Row scaling: Multiplying a row of the identity by a constant gives you a scaling matrix
- Row addition: Adding a multiple of one row to another in the identity gives you an addition matrix
Why This Matters for Inverses
When computing the inverse of matrix A using the augmented matrix method [A | I], every row operation you perform on the left side is like multiplying both sides by an elementary matrix. After converting A to the identity through operations E₁, E₂, ..., Eₖ, you get:
Eₖ · ... · E₂ · E₁ · A = I
This means: A⁻¹ = Eₖ · ... · E₂ · E₁
The inverse is literally the product of all those elementary matrices—which is exactly what appeared on the right side of your augmented matrix!
Key Takeaway: Elementary row operations are matrix multiplications in disguise; finding an inverse through row reduction builds the product of elementary matrices that undo A.