How To Multiply A Matrix By A Vector Step By Step?

Multiplying a matrix by a vector is a core operation in linear algebra used in computer graphics, machine learning, and physics. To do it, you take each row of the matrix and multiply its entries by the corresponding entries in the vector, then add those products together. The result is a new vector whose length matches the number of rows in the matrix.

What Do You Need Before You Start?

Before you multiply, check the dimensions. A matrix has rows and columns, written as rows × columns. A vector is just a matrix with one column.

The rule is strict: the number of columns in the matrix must equal the number of entries in the vector. If the matrix is 3×2 (3 rows, 2 columns), the vector must have 2 entries. If the matrix is 4×3, the vector must have 3 entries.

If the numbers do not match, the multiplication is undefined. There is no workaround. You cannot multiply a 3×2 matrix by a 3-entry vector.

How To Multiply A Matrix By A Vector Step By Step

Here is the exact process. We will use a 2×2 matrix and a 2-entry vector as an example.

Matrix A = [2 3; 1 4] and vector x = [5; 6].

Step 1: Identify the rows of the matrix. Row 1 is [2 3]. Row 2 is [1 4].

Step 2: Multiply the first row by the vector. Take the first entry of the row (2) and multiply it by the first entry of the vector (5). That gives 10. Take the second entry of the row (3) and multiply it by the second entry of the vector (6). That gives 18. Add them: 10 + 18 = 28. This is the first entry of the new vector.

Step 3: Multiply the second row by the vector. Take 1 × 5 = 5. Take 4 × 6 = 24. Add them: 5 + 24 = 29. This is the second entry of the new vector.

Step 4: Write the result. The answer is the vector [28; 29].

That is the entire process. Each row of the matrix produces exactly one number in the output vector.

What Does the Result Mean?

The output vector has the same number of entries as the number of rows in the matrix. A 3×2 matrix times a 2-entry vector gives a 3-entry vector. A 4×4 matrix times a 4-entry vector gives a 4-entry vector.

Think of it as a transformation. The matrix is a rule that takes an input vector and produces an output vector. The output can be longer, shorter, or the same length as the input, depending on the matrix shape.

In computer graphics, this is how 3D points are rotated, scaled, and moved. In machine learning, this is how a layer of a neural network transforms input data. The operation is identical in every field.

What About Larger Matrices?

The process does not change with size. A 3×3 matrix times a 3-entry vector uses the same steps, just with three multiplications per row instead of two.

Consider matrix B = [1 2 3; 4 5 6; 7 8 9] and vector y = [1; 1; 1].

Row 1: (1×1) + (2×1) + (3×1) = 1 + 2 + 3 = 6.
Row 2: (4×1) + (5×1) + (6×1) = 4 + 5 + 6 = 15.
Row 3: (7×1) + (8×1) + (9×1) = 7 + 8 + 9 = 24.

The result is [6; 15; 24].

The pattern is always the same: go across the row, down the vector, multiply pairs, add everything.

Common Mistakes To Avoid

The most frequent error is mixing up rows and columns. You always use the rows of the matrix. Never the columns.

The second most common error is multiplying the wrong entries together. The first entry of the row always pairs with the first entry of the vector. The second with the second. The third with the third. The order never changes.

A third error is forgetting to add the products. Multiplying without adding gives you a list of products, not a single result. The sum is essential.

Finally, do not multiply a vector by a matrix in the same order. Vector times matrix is a completely different operation with different rules. This article covers matrix times vector only.

Why Does Order Matter in Matrix Multiplication?

Matrix multiplication is not commutative. That means the order of the operands changes the result. Matrix times vector is not the same as vector times matrix.

When you multiply a matrix by a vector, the vector must be treated as a column. When you multiply a vector by a matrix, the vector must be treated as a row. The rules for dimensions change completely.

For most practical applications, matrix times vector is the operation you need. It is the standard form used in physics, engineering, and data science.

Can You Multiply a Matrix by a Vector With Fractions or Decimals?

Yes. The arithmetic works exactly the same way. Fractions and decimals follow the same rules of multiplication and addition.

For example, matrix C = [0.5 2; 1.5 3] times vector z = [4; 0.5].

Row 1: (0.5 × 4) + (2 × 0.5) = 2 + 1 = 3.
Row 2: (1.5 × 4) + (3 × 0.5) = 6 + 1.5 = 7.5.

The result is [3; 7.5].

Negative numbers work too. Just keep track of the signs carefully.

What If the Matrix Is Not Square?

Non-square matrices are common. A 2×3 matrix has 2 rows and 3 columns. It can multiply a 3-entry vector and produces a 2-entry vector.

Matrix D = [1 0 2; 3 1 1] times vector w = [2; 1; 3].

Row 1: (1 × 2) + (0 × 1) + (2 × 3) = 2 + 0 + 6 = 8.
Row 2: (3 × 2) + (1 × 1) + (1 × 3) = 6 + 1 + 3 = 10.

The result is [8; 10].

Notice the output has 2 entries because the matrix has 2 rows. The input had 3 entries because the matrix has 3 columns.

When Would You Actually Use This?

This operation appears everywhere in technical work. In computer graphics, every object on a screen is positioned using matrix-vector multiplication. In robotics, joint movements are calculated this way. In economics, input-output models use it to track resources.

In machine learning, a neural network layer multiplies a weight matrix by an input vector to produce an output. This single operation, repeated millions of times, is what allows AI systems to recognize images and process language.

Understanding the manual calculation helps you understand what the computer is doing. It is not magic. It is repeated multiplication and addition.

How Do You Check Your Answer?

You can verify your result by working backward. Take the output vector and see if it makes sense given the input.

A simple check: if the vector contains all zeros, the result must be all zeros. If the matrix is an identity matrix (1s on the diagonal, 0s elsewhere), the result must equal the original vector.

For any other case, redo the calculation row by row. Compare each row’s result independently. If one row is wrong, only that entry changes.

What Is the Difference Between Dot Product and Matrix-Vector Multiplication?

Matrix-vector multiplication is actually a collection of dot products. Each row of the matrix is dotted with the vector to produce one entry of the result.

A dot product takes two equal-length lists of numbers, multiplies them pairwise, and adds the results. That is exactly what you do for each row.

So if you already know how to compute a dot product, you already know how to do matrix-vector multiplication. You just repeat the dot product for every row.

Does the Order of Addition Matter?

No. Addition is commutative and associative. You can add the products in any order and get the same result.

However, you cannot add before multiplying. The multiplication happens first, then the addition. This is a strict rule.

Some people find it helpful to write out each product before summing. This reduces errors, especially with larger matrices.

Why Do Some People Find This Confusing?

The confusion usually comes from trying to memorize the rule instead of understanding it. The rule is simple: row times column, multiply pairs, add.

Another source of confusion is mixing up the dimensions. Always check the column count of the matrix against the entry count of the vector before starting.

Once you understand that each row independently produces one output number, the operation becomes straightforward.

Can You Multiply a Matrix by a Vector in a Spreadsheet?

Yes. Spreadsheet programs like Excel and Google Sheets have built-in functions for this. The MMULT function in Excel performs matrix multiplication.

You select the output range, enter the formula, and press Ctrl+Shift+Enter on Windows or Command+Shift+Enter on Mac. The result fills the selected cells.

Spreadsheets are useful for checking your manual work, especially with larger matrices.