Row Operations on Augmented Matrices

In our last lesson, we learned how to condense a system of linear equations into a single augmented matrix. This was a crucial organizational step, stripping the system down to its numerical essence. Now, we are ready to take the next leap: manipulating this matrix to find the solution to the system.

This lesson introduces the fundamental tools for this process: elementary row operations. You will learn the three "legal" transformations that can be applied to an augmented matrix. Crucially, these operations alter the matrix's appearance without changing the solution of the linear system it represents. Mastering the application of these operations is the central skill that underpins the powerful, algorithmic solution methods we will explore next.

The Three Elementary Row Operations

Think of an augmented matrix as a game board and the elementary row operations as the only moves you are allowed to make. Each row in the matrix represents an equation. Therefore, any operation we perform on a row must correspond to a valid manipulation of that equation. There are exactly three such operations.

Interchange (Swap): Swapping the positions of two rows.
- Notation: $R_{i} \leftrightarrow R_{j}$ (swap row $i$ and row $j$ ).
- Equation equivalent: Reordering the equations in the system, which clearly doesn't affect the final solution.
Scaling (Multiplication): Multiplying all entries in a row by a non-zero constant.
- Notation: $k R_{i} \to R_{i}$ (replace row $i$ with $k$ times row $i$ ).
- Equation equivalent: Multiplying both sides of an equation by a non-zero constant.
Replacement (Addition): Adding a multiple of one row to another row.
- Notation: $R_{i} + k R_{j} \to R_{i}$ (replace row $i$ with the sum of row $i$ and $k$ times row $j$ ).
- Equation equivalent: Adding one equation (or a multiple of it) to another.

The following image provides a concise visual summary of these three operations.

This image illustrates the three types of elementary row operations: interchanging two rows, multiplying a row by a non-zero scalar, and adding a multiple of one row to another.

To see these operations and their notation in action, the following video provides a very clear, step-by-step demonstration.

How To Perform Elementary Row Operations Using Matrices

The video "How To Perform Elementary Row Operations Using Matrices" by The Organic Chemistry Tutor breaks down each of the three operations with simple, clear examples.

Please watch the following segments: Scaling a row: Watch from the beginning to the end of the first example. Swapping two rows: Watch the explanation from the second example. Adding a multiple of one row to another: This is the most complex operation. Watch the detailed walk-through from this segment. Pay close attention to the notation and how the new row is calculated.

The resource below provides a formal definition and examples for each operation. It's a good text to solidify your understanding.

Algebra - Augmented Matrices

This reading from Paul's Online Math Notes defines the elementary row operations and provides a clear example for each one using a 3x3 augmented matrix.

Start at the heading "elementary row operations" and read through the three definitions and their examples. Notice the notation used, such as R_1 \leftrightarrow R_3 for a row swap.

Applying Row Operations to Solve a System

The goal of these operations is not to manipulate the matrix randomly, but to strategically simplify it into a form where the solution is obvious. A matrix that is transformed through these operations is called row equivalent to the original. Since the operations are "legal," row equivalent augmented matrices represent systems with the exact same solution set.

Our ultimate goal is to transform the coefficient part of the matrix (the part to the left of the vertical line) into an identity matrix. If we can achieve this, the solution will appear on the right side.
For a 2x2 system, we want to transform
$[a d b e c f] into [1001 x_{0} y_{0}]$
This final matrix represents the equations $1 x + 0 y = x_{0}$ and $0 x + 1 y = y_{0}$ , which simply means $x = x_{0}$ and $y = y_{0}$ .

Let's see this in practice. The following reading walks through a complete example for a 2x2 system.

Algebra - Augmented Matrices

This section of the same article demonstrates how to use the row operations to solve a 2x2 system. It shows the strategic thinking behind choosing which operation to apply at each step.

Read the section from Example 2. Follow the logic of how they use row operations to create a '1' in the top-left, then a '0' below it, and so on, until the matrix is in the desired final form.

The process for larger systems is identical, just with more steps. The key is to be systematic. A common strategy is to work column by column from left to right, creating a '1' in the diagonal position (the "leading 1") and then using it to create '0's in all other positions in that column.

The following video demonstrates this strategy on a 3x3 system. The presenter uses a specific order to create the zeros, which is a preview of the formal Gaussian elimination algorithm we will study next. For now, focus on how each operation is chosen and applied to achieve a specific goal (like creating a zero in a certain position).

Elementary Row Operations - Solve Using an Augmented Matrix

In "Elementary Row Operations - Solve Using an Augmented Matrix," Mario's Math Tutoring provides a complete, worked example of solving a 3x3 system. He emphasizes a systematic approach to avoid undoing previous work.

Watch from the start of the process. The presenter explains a "cheat code" or order for getting the zeros. Follow his application of row operations to clear out the first column (getting zeros below the first leading 1) and then to simplify the second and third columns. Notice how combining a row with another that has a leading zero can be used to preserve work you've already done.

The most important operation is replacement ( $R_{i} + k R_{j} \to R_{i}$ ), as it's the primary tool for creating zeros. The general tactic is: to create a zero at a certain position, use a multiple of another row that has a non-zero entry (ideally a '1') in the same column. For instance, to zero out the entry in row 2, column 1 ( $a_{21}$ ), you might use a multiple of row 1, which has entry $a_{11}$ : $R_{2} - \frac{a _{21}}{a _{11}} R_{1} \to R_{2}$ .

Conclusion

In this lesson, we have introduced the three elementary row operations, the fundamental building blocks for solving linear systems with matrices. These operations provide a powerful way to transform a complicated system into an equivalent, simpler one from which the solution can be easily read.

Here are the main takeaways:

The three elementary row operations are swapping rows, scaling a row by a non-zero constant, and adding a multiple of one row to another.
These operations produce a row equivalent matrix, which represents a linear system with the same solution set as the original.
The strategic goal is to apply these operations to transform the augmented matrix into a simpler form (ideally with an identity matrix on the left), which reveals the system's solution.

We have now learned what the moves are and have seen them applied. In our next lesson, we will formalize the strategy. We will learn a step-by-step algorithm called Gaussian elimination, which uses these row operations to systematically reduce any matrix to a standard form called row echelon form, bringing order and predictability to the solution process.

Recommended resources

The Three Elementary Row Operations

Applying Row Operations to Solve a System

Conclusion