Course contentsShow
← Mathematics
Lesson 2127 of 3,12144. Linear Algebra — Matrices and SystemsFree lesson

Computing Rank via Row Reduction

Finding the rank of a matrix by counting pivot positions in its row echelon form.

Computing Rank via Row Reduction

You'll learn: How to find the rank of any matrix by counting the pivot positions after performing row reduction.

The Core Idea

Remember that the rank of a matrix is the dimension of its column space—essentially, the number of linearly independent columns. But how do we actually compute it?

The answer is beautifully simple: perform row reduction to get the matrix into row echelon form (REF), then count the pivot positions. Each pivot represents one dimension of independent information.

Why This Works

Row operations don't change the fundamental relationships between columns. They preserve linear independence. When you reduce a matrix to REF, the pivot positions reveal exactly which columns contain truly independent information—the rest are redundant combinations.

Think of it like decluttering: row reduction strips away all the redundancy, and the pivots are what remains—the "essential" rows that can't be eliminated.

The Process

  1. Start with your matrix (any size)
  2. Apply elementary row operations to convert it to row echelon form
  3. Count the number of pivot positions (leading entries in non-zero rows)
  4. That count is the rank

You can actually use either REF or RREF—both will give you the same number of pivots. RREF is often clearer because pivots are exactly 1 and easier to spot.

Example Thinking

If you reduce a 4×5 matrix and end up with 3 non-zero rows (each with a pivot), the rank is 3. This tells you that only 3 of the original columns were truly independent; the other 2 were linear combinations of those three.

Key Takeaway: To compute the rank of a matrix, row-reduce it to row echelon form and count the pivot positions—this number equals the rank.