What a matrix is, its dimensions and elements, when two matrices are equal, adding and subtracting element by element, scalar multiplication, combining the operations, and organising real data in matrix form.
Subject: Algebra 2 · 69 slides · symbolic lesson
Open the interactive version of this deck
Title
Algebra 2 · Lesson 3.5
A matrix is a table you can do arithmetic on, one position at a time
Objectives
Chapter 3 has been about systems of equations, and every method so far has made you copy the letters out again on every line. This lesson introduces the object that eventually lets you stop: a matrix, which is a rectangular arrangement of numbers in rows and columns.
Nothing here is difficult. The arithmetic is addition, subtraction and multiplication of ordinary numbers, done one position at a time. What is new is the bookkeeping, and the bookkeeping is where marks are lost, so the lesson is careful about it.
The last two objectives are the ones worth holding on to. A matrix is not decoration; it is a table that means something, and every operation in this lesson has a plain-English reading in terms of the data the table holds.
One habit will save you throughout. Before you do anything to two matrices, look at their shapes. Almost every question in this lesson either works immediately or is undefined, and the shapes tell you which before you write a single number.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 187 — Before, Now, Why and the Key Vocabulary
Warm-up
Start where the textbook starts, with data rather than notation. A company makes DVD racks in two sizes, small and large, from three woods: walnut, pine and cherry. Last month's sales are in the table below.
Figure (svg): A two-way table of last month's DVD rack sales: the rows are labelled walnut, pine and cherry, the columns small and large, and the six numbers are 125 and 100, 278 and 251, 225 and 270.
Discussion prompt
Read three numbers out of the table: how many small pine racks sold, how many large cherry racks sold, and how many walnut racks sold altogether. Then say in one sentence what you had to know in order to find each one.
Notice how little of the table is actually numbers. Six of the boxes hold data; the rest are words that tell you where you are. Those words are labels, and the arithmetic never touches them.
That observation is the doorway into this lesson. If the labels never take part in the calculation, they can be moved outside, and what is left is six numbers in a fixed arrangement. That arrangement is a matrix.
Keep this table in your head for the next hour. Every operation you meet will be tried out on it, and each one will turn out to answer a question the company might genuinely ask.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 189 — the DVD rack sales data from Example 3
Section
Part 1
Concept
Here is the same table with the labels moved outside the brackets. Nothing about the data has changed; all that has happened is that the part the arithmetic will use has been fenced off from the part it will not.
Figure (svg): The same six sales numbers enclosed in square brackets, with the words walnut, pine and cherry written outside to the left and small and large outside above, and the note that the matrix is 3 by 2.
matrix — A rectangular arrangement of numbers in rows and columns. The numbers inside it are its elements, and a capital letter such as A is used to name the whole arrangement.
Calling the whole arrangement A is the single most useful piece of notation in the chapter. From now on one letter stands for six numbers, and later in Chapter 3 that same letter will stand for an entire system of equations.
The brackets are not decoration either. They are what says these numbers travel together. Write the six numbers without brackets and you have a list; write them inside brackets in a fixed grid and you have an object with a shape.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 187 — the definition of a matrix
Concept
The size of a matrix is described by two numbers, and the order they are written in is fixed. Count the horizontal lines of numbers first and the vertical ones second.
dimensions — The size of a matrix, written as the number of rows by the number of columns, in that order. A matrix with m rows and n columns is described as m by n.
\[ A = \begin{bmatrix}4 & -1 & 5 \\ 0 & 6 & 3\end{bmatrix} \qquad \text{dimensions } 2 \times 3 \]
Figure (svg): The matrix A with rows 4, minus 1, 5 and 0, 6, 3, with a bracket marking two rows, a bracket marking three columns, and an arrow pointing at the 5 in the first row and third column.
Two rows and three columns, read aloud as two by three. Getting the order the wrong way round is not a small slip: a 2 by 3 matrix and a 3 by 2 matrix are different objects, and in the next lesson one of them will be multipliable where the other is not.
A memory hook that works: rows run across, like the rows of seats in a cinema, and you say which row you are in before you say which seat. Matrices keep the same order.
One consequence is worth noticing straight away. A matrix is a full rectangle, so every row holds the same number of entries, and the total count of elements is always the number of rows multiplied by the number of columns. Here that is six.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 187 — dimensions and elements
Concept
Nothing in the definition says a matrix has to be tall or wide. A single horizontal line of numbers is a matrix, so is a single vertical line, and both come up constantly once data is being organised.
\[ \begin{bmatrix}2 & -6 & 1\end{bmatrix} \quad 1 \times 3 \qquad \begin{bmatrix}7 \\ -3\end{bmatrix} \quad 2 \times 1 \qquad \begin{bmatrix}4 & -1 \\ 0 & 6\end{bmatrix} \quad 2 \times 2 \]
The first is one row of three entries, so it is 1 by 3. The second is two rows of one entry each, so it is 2 by 1. The third has as many rows as columns, which makes it square, and square matrices are the ones Lessons 3.7 and 3.8 are entirely about.
The row and the column are worth separating carefully, because they are the pair most often confused. Both of them are just a list of numbers as far as the values go, and they are different matrices because they are laid out along different directions.
You have already met both of them without the names. A single row of a data table is a row matrix, and the list of one product's sales down the months is a column matrix, and later in this chapter a whole system of equations gets written as a column of constants.
Watch out for one piece of everyday language. Saying a matrix is 3 by 1 does not mean it is small or simple; it means it has three rows and one column, and it can still be added only to another 3 by 1 matrix.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 187 — dimensions written as m by n
Notation
Three pieces of notation arrive at once, and they are easy to keep straight if you decode them now rather than meeting them cold in an exercise.
Annotate
On: \( A = \begin{bmatrix}a_{11} & a_{12} & a_{13} \\ a_{21} & a_{22} & a_{23}\end{bmatrix} \)
The lowered numbers are subscripts. They look a little like exponents at a glance, and the difference matters: a raised number would tell you to multiply something by itself, whereas a lowered one is only a label saying which entry is meant.
You will meet this address notation again in Lesson 3.6 and heavily in Lesson 3.7, where the whole method depends on being able to say which row and which column you are working in. Learning it properly now costs a minute and saves an afternoon.
If the double subscript ever confuses you, read it out in words instead of symbols. Saying the entry in row two, column three is slower to write and impossible to misread.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 187 — elements and the element in row 1, column 3
Worked example
Take the book's own matrix and answer the two questions that are asked about every matrix you will ever meet: what size is it, and what sits at a given address?
\[ A = \begin{bmatrix}4 & -1 & 5 \\ 0 & 6 & 3\end{bmatrix} \]
Count the rows
Why: a row is a horizontal line of entries, and there are two of them here.
\[ \text{rows} = 2 \]
Count the columns
Why: a column is a vertical line of entries, and there are three.
\[ \text{columns} = 3 \]
Write the dimensions in the fixed order
Why: rows first, then columns, so this is a 2 by 3 matrix and not a 3 by 2 one.
\[ 2 \times 3 \]
Now for the address. Work down to the row you want before you work across to the column, in the same order the notation is written.
Go down to row 2, then across to column 2
Why: row 2 holds 0, 6 and 3, and the second of those is 6.
\[ a_{22} = 6 \]
Check by counting the elements two ways
Why: two rows of three entries should give six elements, and counting the numbers straight off the page also gives six, so nothing has been miscounted or left out.
\[ 2 \times 3 = 6 \quad \text{elements} \]
That count is a genuinely useful check. Whenever a matrix is copied out by hand, an element goes missing far more often than an element goes wrong, and the count catches it.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 187 — matrix A and its element in row 1, column 3
Picture it
Before any arithmetic, get comfortable with the idea that a position exists independently of what is sitting in it. That is what makes element-by-element operations possible at all.
Figure (svg): A two by three grid in which every cell is labelled with its own address, row 1 column 1 through row 2 column 3, with an arrow down the side marked rows and an arrow across the top marked columns.
Each cell in the picture is labelled with its address rather than a value. If you handed somebody an empty grid like this and read out six numbers with their addresses, they could reconstruct the matrix exactly.
The two arrows are the reading order made visible. You travel down to the row, then across to the column, which is why the dimensions and the addresses are both written rows first.
This also explains something that will come up again in Part 2. When two matrices are added, what is really being paired off is not numbers but positions, and positions can only be paired if the two grids have the same shape.
Fill the middle
Practise the address notation on the matrix you already know, with rows 4, negative 1, 5 and then 0, 6, 3. Work down to the row first and across to the column second, and type just the number that lives there.
Fill in the blanks
a_-1 = 3, \qquad a_0 = ___, \qquad a____ = ___
Why: Row 1 is 4, negative 1, 5, so the second entry of it is negative 1. Row 2 is 0, 6, 3, so its third entry is 3 and its first entry is 0. Reading the subscripts in the other order would have asked for row 2 column 1 instead of row 1 column 2, which is a different number entirely.
The middle one is the trap. Reading the pair backwards turns the entry in row 1, column 2 into the entry in row 2, column 1, and here those are negative 1 and 0, so the mistake is invisible unless you check the order deliberately.
Notice also that there is no such element as the one in row 3. The matrix has only two rows, so an address with a 3 in the first slot is not a wrong answer, it is a request for something that does not exist.
If you found yourself counting on your fingers, that is exactly right for now. Speed comes later, and an address read slowly and correctly beats one read quickly and backwards.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 187 — naming the element in a given row and column
Prediction
Commit to a number before you reason it out. Most wrong answers here come from treating the two dimensions as things to combine rather than as a description of a full rectangle.
Predict first
How many elements does a matrix with 4 rows and 7 columns have?
Correct: 28
Why: A matrix is a complete rectangle, so every row has the same length. Four rows of seven entries is 28 entries, which is the number of rows multiplied by the number of columns. Adding the dimensions would give 11, and writing them side by side would give 47; neither counts anything.
\[ 4 \times 7 = 28 \]
The reason this works is that no position is ever allowed to be empty. A table of data with a gap in it is a perfectly ordinary thing, but it is not a matrix, and you would have to put a zero or some other agreed value in the gap before treating it as one.
Turn the fact round and it becomes a useful check. If somebody hands you 28 numbers and says they form a matrix with four rows, you already know each row must hold seven, so you can lay it out without being told.
The same count also tells you when two matrices cannot possibly be equal or addable. Different element counts means different shapes, although matching element counts is not enough on its own, as the very next slide shows.
Concept
Equality for matrices is stricter than equality for lists of numbers, and the extra strictness is the whole point of the definition.
equal matrices — Two matrices are equal when their dimensions are the same and the elements in corresponding positions are equal.
Figure (svg): Two two by two matrices side by side, the first with rows 1, 2 and 3, 4 and the second with rows 1, 3 and 2, 4. The two positions that agree are ringed in green and the two that disagree are ringed in red.
The two matrices in the picture are built from the same four numbers and have the same shape, and they are still not equal, because the entry in row 1, column 2 is 2 in one and 3 in the other.
\[ \begin{bmatrix}1 & 2 \\ 3 & 4\end{bmatrix} \ne \begin{bmatrix}1 & 3 \\ 2 & 4\end{bmatrix} \]
Compare this with ordinary numbers for a moment. The digits 1 and 2 make both 12 and 21, and nobody would call those the same number, because position carries meaning. A matrix works the same way, only in two directions at once.
This definition is doing more work than it looks. Part 5 of this lesson solves equations by setting two matrices equal to each other, and the reason that works at all is that equality forces every single position to agree.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 187 — the definition of equal matrices
Sorting
Test the definition rather than your instinct. For each pair, ask the two questions in order: do the dimensions match, and then does every corresponding position hold the same number?
Sort into buckets
Sort each pair of matrices by whether the two are equal.
The pair of a column against a row is the one people argue about. Both hold a 7 and a negative 3, and they are still different matrices, because one is 2 by 1 and the other is 1 by 2.
The pair that differs in a single entry is worth a second look too. Five of the six positions agree, which feels like almost equal, but equality has no almost: one disagreement anywhere settles it.
Notice how mechanical the test is once you trust it. Dimensions first, because if those fail you can stop; only then do you walk through the positions in order.
Trap
A matrix has two rows and three columns, and the size is written down as:
\[ 3 \times 2 \]
It looks harmless. The two right numbers are there, and in ordinary arithmetic the order of a product makes no difference at all.
That instinct from multiplication is exactly what causes the slip. Here the two numbers are not being multiplied together; they are a description, and a description read backwards describes something else.
Rows first, columns second, with no exceptions:
\[ 2 \times 3 \]
A 3 by 2 matrix is a genuinely different object: it would have three horizontal lines of two entries rather than two horizontal lines of three.
Both have six elements, which is why the error survives a quick sanity check. Counting elements can never tell the two apart, and only counting rows can.
In Lesson 3.6 this stops being a matter of tidiness. Whether two matrices can be multiplied depends on their dimensions in a way that is not symmetric, so a reversed pair of numbers there gives a wrong answer rather than an untidy one.
Section
Part 2
Concept
Here is the entire rule for adding two matrices, and it is as simple as it sounds: pair off the positions, add the two numbers in each pair, and write the answer in that position of a new matrix.
Figure (svg): Two two by two matrices added, with the middle matrix showing the four sums written out, 3 plus minus 1, 0 plus 4, minus 5 plus 2 and minus 1 plus 0, and the last matrix showing the results 2, 4, minus 3 and minus 1.
Look at the third matrix in the picture, the one still holding sums rather than answers. That is what matrix addition really is, and the final matrix is only that one after the arithmetic has been done.
Nothing leaks between positions. The entry in row 1, column 1 of the answer is built from the two entries in row 1, column 1 and from nothing else, which is why you can work through a large matrix sum in any order you like without keeping track of anything.
Subtraction is the same sentence with one word changed: subtract the numbers in corresponding positions. Everything said about addition on this slide applies to it unchanged.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 187 — Key Concept: Adding and Subtracting Matrices
Notation
The textbook states the rule with letters so that it covers every 2 by 2 case at once. Decode it once here and you will not have to read it again.
Annotate
On: \( \begin{bmatrix}a & b \\ c & d\end{bmatrix} + \begin{bmatrix}e & f \\ g & h\end{bmatrix} = \begin{bmatrix}a+e & b+f \\ c+g & d+h\end{bmatrix} \)
The subtraction rule in the book sits directly underneath and differs by a single symbol in each of the four positions. Once you have understood one, the other is not a second thing to learn.
Statements with letters like this one deserve a habit: read the rule, then immediately test it on numbers you can check. Putting 3, 0, negative 5 and negative 1 in for the first four letters gives exactly the sum you saw on the previous slide.
It is worth noticing what the rule quietly assumes. It is written with two matrices of the same shape, and there is no version of it for mismatched shapes, which is the subject of the next few slides.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 187 — the Adding and Subtracting Matrices box
Worked example
Start with the book's first example. Before adding anything, do what the margin note beside it tells you to do and check the two shapes.
\[ \begin{bmatrix}3 & 0 \\ -5 & -1\end{bmatrix} + \begin{bmatrix}-1 & 4 \\ 2 & 0\end{bmatrix} \]
Check the dimensions agree
Why: both matrices have two rows and two columns, so every position in one has a partner in the other and the sum exists.
\[ 2 \times 2 \quad \text{and} \quad 2 \times 2 \]
Write the sum position by position
Why: keep the additions unresolved for one line, so that you can see the pairing before the arithmetic hides it.
\[ \begin{bmatrix}3 + (-1) & 0 + 4 \\ -5 + 2 & -1 + 0\end{bmatrix} \]
Writing the brackets round the negative numbers is worth the extra ink. Three plus negative one is unambiguous; three minus one looks like a subtraction problem you were not given.
Do the four additions
Why: each one is an ordinary sum of two integers, and the signs are handled exactly as they were in Chapter 1.
\[ \begin{bmatrix}2 & 4 \\ -3 & -1\end{bmatrix} \]
Check by subtracting one of the originals back off
Why: taking the second matrix away from the answer should return the first: 2 minus negative 1 is 3, 4 minus 4 is 0, negative 3 minus 2 is negative 5, and negative 1 minus 0 is negative 1, which is the first matrix exactly.
\[ \begin{bmatrix}2 & 4 \\ -3 & -1\end{bmatrix} - \begin{bmatrix}-1 & 4 \\ 2 & 0\end{bmatrix} = \begin{bmatrix}3 & 0 \\ -5 & -1\end{bmatrix} \]
That check costs one line and tests all four positions at once, which makes it much better value than re-adding the same four pairs and repeating any slip you made the first time.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 187 — Example 1(a)
Worked example
The second half of the book's Example 1 is a subtraction of two 3 by 2 matrices, and it is deliberately loaded with negative numbers. This is where most of the real difficulty in the lesson lives.
\[ \begin{bmatrix}7 & 4 \\ 0 & -2 \\ -1 & 6\end{bmatrix} - \begin{bmatrix}-2 & 5 \\ 3 & -10 \\ -3 & 1\end{bmatrix} \]
Check the dimensions agree
Why: three rows and two columns in both, so the difference is defined and will itself be 3 by 2.
\[ 3 \times 2 \quad \text{and} \quad 3 \times 2 \]
Write the differences position by position
Why: put every subtracted negative in brackets, because minus negative two is the exact place this example is designed to catch you.
\[ \begin{bmatrix}7 - (-2) & 4 - 5 \\ 0 - 3 & -2 - (-10) \\ -1 - (-3) & 6 - 1\end{bmatrix} \]
Three of those six are a subtraction of a negative number, and each one turns into an addition. Seven minus negative two is nine, not five, and that single move is the difference between this example being easy and being a mess.
Work out the six entries
Why: top row nine and negative one, middle row negative three and eight, bottom row two and five.
\[ \begin{bmatrix}9 & -1 \\ -3 & 8 \\ 2 & 5\end{bmatrix} \]
Look at the answer against the two inputs for a moment. The entry that grew the most is the one where a negative ten was taken away, which is exactly what you would expect: removing a debt makes a number larger.
Check by adding the subtracted matrix back on
Why: the answer plus the second matrix should return the first: nine plus negative two is seven, negative one plus five is four, negative three plus three is zero, eight plus negative ten is negative two, two plus negative three is negative one, and five plus one is six.
\[ \begin{bmatrix}9 & -1 \\ -3 & 8 \\ 2 & 5\end{bmatrix} + \begin{bmatrix}-2 & 5 \\ 3 & -10 \\ -3 & 1\end{bmatrix} = \begin{bmatrix}7 & 4 \\ 0 & -2 \\ -1 & 6\end{bmatrix} \]
All six positions came back, so every sign was handled correctly. Reversing the operation is the cheapest complete check there is, and with matrices it tests every entry at once.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 188 — Example 1(b)
Concept
The textbook states subtraction in its own box, directly under the addition box, and the two statements differ by a single symbol in each position. That is worth looking at rather than skimming past.
\[ \begin{bmatrix}a & b \\ c & d\end{bmatrix} - \begin{bmatrix}e & f \\ g & h\end{bmatrix} = \begin{bmatrix}a-e & b-f \\ c-g & d-h\end{bmatrix} \]
Every remark made about addition carries over unchanged. The positions are paired the same way, no position affects any other, the dimensions must match, and the answer has the same shape as the two inputs.
One thing does change, and it is the thing to be careful about. Addition does not care which matrix you write first, because in each position the two numbers can be added in either order. Subtraction cares very much, because in each position the order decides the sign.
\[ \begin{bmatrix}a & b \\ c & d\end{bmatrix} - \begin{bmatrix}e & f \\ g & h\end{bmatrix} \ne \begin{bmatrix}e & f \\ g & h\end{bmatrix} - \begin{bmatrix}a & b \\ c & d\end{bmatrix} \]
The two differences are opposites of each other: every entry of one is the negative of the matching entry of the other. Part 6 uses that directly, where this month minus last month records rises and last month minus this month records the same changes with the signs reversed.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 187 — the Subtracting Matrices box
Concept
Adding pairs off positions, so two matrices can only be added when every position in one has a partner in the other. That is a statement about shape, and nothing about the values can rescue it.
Figure (svg): A two by three grid beside a three by two grid, with a plus sign and a crossed circle between them, showing that positions in one have no partner in the other.
The two grids in the picture both hold six numbers, and they still cannot be added. The 2 by 3 grid has a position in row 1, column 3; the 3 by 2 grid has no column 3 at all, so there is nothing to pair that position with.
The book puts this in an Avoid Errors note for a reason. The mistake is rarely made deliberately. It happens when somebody starts adding, gets through four entries comfortably, and only discovers the problem when the fifth has no partner.
So make the shape check the first line of your working rather than a thing you discover. Write down both sets of dimensions, compare them, and only then start adding. An answer of undefined is a complete and correct answer when the shapes disagree.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 187 — the Avoid Errors note
Discrimination
Six expressions, and the only question is whether the operation is possible at all. Do not work any of them out; decide on the shapes alone, which is the whole skill being practised.
Sort into buckets
Sort each expression by whether it is defined.
The single row against the single column is the sharpest case. Four numbers each, and no position in one corresponds to any position in the other, because a row and a column are laid out along different directions.
Notice that subtraction never behaved differently from addition anywhere in that list. The two operations have exactly the same requirement, since both of them work by pairing positions.
One more thing falls out of the exercise for free. Whenever an operation is defined, you already know the shape of the answer before doing any arithmetic, because it is the shared shape of the two inputs.
Trap
Two matrices are offered for addition, with sizes:
\[ (2 \times 3) + (3 \times 2) \]
Each holds six numbers, so it is tempting to add them six numbers at a time and write down a result.
Something can always be written. Pair the entries off in reading order and six sums appear, and the page looks exactly like a correct solution.
There is no such sum. The right answer is:
\[ \text{undefined} \]
The pairing that made those six sums was invented. Row 1, column 3 of the first matrix has no partner, because the second matrix has no third column.
The giveaway is that the shape of any answer would be a guess. Genuine matrix addition hands you the answer's dimensions automatically, and here it cannot.
Write both sets of dimensions down before you add anything. It is one line, it takes three seconds, and it turns this trap into something you never even step towards.
Worked example
The book's first checkpoint. Same method as Example 1, one column wider, and with a mix of signs in both matrices so that no position is free.
\[ \begin{bmatrix}-2 & 5 & 11 \\ 4 & -6 & 8\end{bmatrix} + \begin{bmatrix}-3 & 1 & -5 \\ -2 & -8 & 4\end{bmatrix} \]
Check the dimensions agree
Why: two rows and three columns in each, so the sum exists and will be 2 by 3.
\[ 2 \times 3 \quad \text{and} \quad 2 \times 3 \]
Write the six sums position by position
Why: work along the first row and then along the second, in the same order you read.
\[ \begin{bmatrix}-2 + (-3) & 5 + 1 & 11 + (-5) \\ 4 + (-2) & -6 + (-8) & 8 + 4\end{bmatrix} \]
Two of these are a negative added to a negative, which makes the answer more negative rather than less. Negative 6 plus negative 8 is negative 14, and getting that to negative 2 is the commonest slip in this checkpoint.
Do the arithmetic
Why: top row negative five, six and six, bottom row two, negative fourteen and twelve.
\[ \begin{bmatrix}-5 & 6 & 6 \\ 2 & -14 & 12\end{bmatrix} \]
Check by subtracting the second matrix back off
Why: negative five minus negative three is negative two, six minus one is five, six minus negative five is eleven, two minus negative two is four, negative fourteen minus negative eight is negative six, and twelve minus four is eight, which rebuilds the first matrix.
\[ \begin{bmatrix}-5 & 6 & 6 \\ 2 & -14 & 12\end{bmatrix} - \begin{bmatrix}-3 & 1 & -5 \\ -2 & -8 & 4\end{bmatrix} = \begin{bmatrix}-2 & 5 & 11 \\ 4 & -6 & 8\end{bmatrix} \]
Notice that the check is itself a subtraction of matrices, so doing it also gives you a free rehearsal of the operation from the previous example.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 188 — Guided Practice 1
Comparison
This checkpoint subtracts a 3 by 2 matrix from another. The first matrix has rows negative 4 and 0, then 7 and negative 2, then negative 3 and 1; the second has rows 2 and 2, then negative 3 and 0, then 5 and negative 14.
\[ \begin{bmatrix}-4 & 0 \\ 7 & -2 \\ -3 & 1\end{bmatrix} - \begin{bmatrix}2 & 2 \\ -3 & 0 \\ 5 & -14\end{bmatrix} \]
Three of the six entries are missing below. Work each one out on its own, subtracting the second matrix's entry from the first matrix's entry in the same position, and type the number.
Comparison matrix
| position | column 1 | column 2 |
|---|---|---|
| row 1 | -6 | −2 |
| row 2 | 10 | -2 |
| row 3 | −8 | 15 |
The entry in the last row and last column is the one to take slowly. One minus negative fourteen is one plus fourteen, which is fifteen, and it is the largest number anywhere in the problem even though neither input entry was large.
The first row is a different kind of care: negative four minus two is negative six, not negative two. Subtracting a positive number from a negative one drives it further down, and reading it as though the signs cancelled is a very easy slip to make quickly.
Once every box is filled, read the completed grid as a matrix and give its dimensions. It is 3 by 2, the same shape as both inputs, which is what subtraction always does.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 188 — Guided Practice 2
Error analysis
This is the error-analysis exercise the textbook sets straight after the lesson. Somebody has added two single-column matrices and produced something with two columns. Find the moment it went wrong.
Annotate
On: \( \begin{bmatrix}9 \\ -5\end{bmatrix} + \begin{bmatrix}4.1 \\ 3.8\end{bmatrix} = \begin{bmatrix}9 & 4.1 \\ -5 & 3.8\end{bmatrix} \)
\[ \begin{bmatrix}9 \\ -5\end{bmatrix} + \begin{bmatrix}4.1 \\ 3.8\end{bmatrix} = \begin{bmatrix}13.1 \\ -1.2\end{bmatrix} \]
What makes this error worth studying is that it is not an arithmetic slip. Every number in the wrong answer is a number from the question, correctly copied, and the mistake is entirely about what the operation means.
The shape test catches it instantly, which is why it is such a good habit. Say the dimensions of both inputs out loud, say the dimensions the answer must therefore have, and then compare that with what you have written.
Decimals appearing in the entries changes nothing at all here. A matrix element can be any real number, and negative 5 plus 3.8 is handled exactly the way it would be outside a matrix.
One last observation about the wrong answer. It has four entries where the inputs had two each, so it has invented no information but rearranged it. That is the signature of a placing-side-by-side error rather than a calculation error.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 190 — Exercise 3, Error Analysis
Prediction
Before any numbers, a question about shape alone. Two matrices, each with three rows and four columns, are added together.
Predict first
Two matrices each have 3 rows and 4 columns. What are the dimensions of their sum?
Correct: 3 by 4
Why: Every position of the sum is built from one position of each input, so the sum has exactly one entry for each position the inputs share. That gives the same 3 rows and 4 columns, whatever the numbers happen to be.
\[ (3 \times 4) + (3 \times 4) = 3 \times 4 \]
The wrong options are each a real habit misfiring. Swapping the two numbers is the rows-and-columns slip from Part 1, and multiplying the dimensions together is a memory of counting elements rather than describing a shape.
The final option is worth dismissing properly rather than just rejecting. The values inside a matrix never affect its dimensions, because the dimensions describe the arrangement and not the contents. You could replace all twelve entries with zeros and the shape would be untouched.
Knowing the answer's shape in advance is more useful than it sounds. It tells you how many entries to write, so if your working produces eleven or thirteen numbers you know to stop and look for the missing or repeated position.
Section
Part 3
Concept
In matrix algebra an ordinary real number is called a scalar, to distinguish it from a matrix. Multiplying a matrix by a scalar means multiplying every single element by it.
scalar multiplication — Multiplying a matrix by a real number, called a scalar, by multiplying each element of the matrix by that number. The result has the same dimensions as the original.
Figure (svg): The scalar minus 2 inside a circle with six arrows fanning out to the six entries of a three by two matrix, and the resulting matrix drawn beside it.
Count the arrows in the picture against the entries of the matrix. There are six of each, and that correspondence is the whole rule. A scalar that reaches five entries out of six has not been applied.
There is no shape requirement here, which is a relief after Part 2. A scalar can multiply a matrix of any dimensions at all, because a single number has no shape to clash with.
Nor does the shape change. Six entries go in and six entries come out, in the same arrangement, so scalar multiplication moves the values without disturbing the structure.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 188 — Scalar Multiplication
Intuition
Seeing the rule stated is one thing; seeing what it does to the numbers is another. Draw each entry as a bar measured from zero and the effect of the scalar becomes a single, uniform change.
Figure (svg): Four pairs of bars measured from a zero line. In each pair the left bar is an original entry, 4, minus 1, 2 and 7, and the right bar is that entry multiplied by minus 2, giving minus 8, 2, minus 4 and minus 14.
Every bar has doubled in length, because the scalar has size 2, and every bar has crossed to the other side of zero, because the scalar is negative. Both of those happen to all four entries and not to a chosen few.
This is where the name earns itself. Scaling is what the number does: it changes the scale of everything in the matrix uniformly, the way a map scale changes every distance on the map by the same factor.
The picture also makes a prediction you can test. Multiply by a scalar between zero and one and every bar shrinks toward zero without changing side, which is exactly what halving the total sales matrix does in Example 3 later on.
And multiplying by zero would flatten every bar at once, giving a matrix of the right shape with nothing but zeros in it. That matrix has a name in later courses, and you can already see where it comes from.
Worked example
The book's Example 2 begins with a negative scalar on purpose, because handling the sign in six places at once is the part that needs practice.
\[ -2\begin{bmatrix}4 & -1 \\ 1 & 0 \\ 2 & 7\end{bmatrix} \]
Write the scalar against every entry
Why: six entries means six separate products, and writing them out before evaluating them is what stops one being skipped.
\[ \begin{bmatrix}-2(4) & -2(-1) \\ -2(1) & -2(0) \\ -2(2) & -2(7)\end{bmatrix} \]
Notice there is no shape check to do this time. A scalar has no dimensions to clash with, so the only thing that could go wrong is missing an entry, which is why every entry is written out first.
Work out the six products
Why: watch the two sign rules: negative two times negative one is positive two, and negative two times zero is zero rather than negative zero.
\[ \begin{bmatrix}-8 & 2 \\ -2 & 0 \\ -4 & -14\end{bmatrix} \]
Five of the six answers came out negative, and the one exception is the entry that started negative. That pattern is a quick plausibility check whenever the scalar is negative.
Check by dividing the answer by the scalar
Why: negative eight over negative two is four, two over negative two is negative one, negative two over negative two is one, zero over negative two is zero, negative four over negative two is two, and negative fourteen over negative two is seven, which is the original matrix.
\[ -\tfrac12\begin{bmatrix}-8 & 2 \\ -2 & 0 \\ -4 & -14\end{bmatrix} = \begin{bmatrix}4 & -1 \\ 1 & 0 \\ 2 & 7\end{bmatrix} \]
Dividing back through is the natural check for scalar multiplication, and multiplying by negative one half is the same thing written as a scalar, so the check is itself an example of the operation.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 188 — Example 2(a)
Comparison
Practise the rule on a fresh matrix with rows 3 and negative 6, then negative 1 and 0, then 5 and 2. The scalar is 4, so every entry is multiplied by 4 and nothing else happens.
\[ 4\begin{bmatrix}3 & -6 \\ -1 & 0 \\ 5 & 2\end{bmatrix} \]
Three of the six results are missing. Each blank is one product of the scalar with the entry in that position, so work them out one at a time rather than looking for a pattern down a column.
Comparison matrix
| position | column 1 | column 2 |
|---|---|---|
| row 1 | 12 | -24 |
| row 2 | -4 | 0 |
| row 3 | 20 | 8 |
The entry that was zero stays zero, because four times zero is zero. That is the only entry a scalar cannot change, and spotting it is a small free win in any question of this kind.
The negative entries keep their sign here, because the scalar is positive. Compare that with Example 2, where a negative scalar reversed every sign, and you have the whole story of what the sign of a scalar does.
Check the finished grid by adding it up against the original. Every entry should be exactly four times as far from zero as it was, in the same direction, and the shape should still be 3 by 2.
Worked example
A larger matrix, so nine products rather than six, and a scalar that is both negative and larger than two. The method does not change at all, which is the point of doing it.
\[ -4\begin{bmatrix}2 & -1 & -3 \\ -7 & 6 & 1 \\ -2 & 0 & -5\end{bmatrix} \]
Multiply along the first row
Why: negative four times two is negative eight, times negative one is four, and times negative three is twelve.
\[ \begin{bmatrix}-8 & 4 & 12\end{bmatrix} \]
Multiply along the second row
Why: negative four times negative seven is twenty-eight, times six is negative twenty-four, and times one is negative four.
\[ \begin{bmatrix}28 & -24 & -4\end{bmatrix} \]
Multiply along the third row
Why: negative four times negative two is eight, times zero is zero, and times negative five is twenty.
\[ \begin{bmatrix}8 & 0 & 20\end{bmatrix} \]
Working a row at a time is a good discipline on a matrix this size. It gives you three short calculations with a natural stopping point rather than nine that blur together.
Assemble the answer
Why: put the three rows back into a 3 by 3 matrix, in the order they came out.
\[ \begin{bmatrix}-8 & 4 & 12 \\ 28 & -24 & -4 \\ 8 & 0 & 20\end{bmatrix} \]
Check the signs against the original
Why: every entry that was positive is now negative and every entry that was negative is now positive, except the zero, which cannot change side; and each entry is four times as far from zero as it was.
\[ \text{for example } -4(-7) = 28 \quad \text{and} \quad -4(6) = -24 \]
Two independent checks in one line, one about sign and one about size, and between them they catch almost every mistake this kind of question can produce.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 188 — Guided Practice 3
Prediction
A quick question about shape rather than value. A 5 by 2 matrix is multiplied by the scalar 7.
Predict first
A 5 by 2 matrix is multiplied by the scalar 7. What happens to its dimensions?
Correct: Nothing at all
Why: Scalar multiplication replaces each entry by a new number in the same position. No position is created and none is destroyed, so a 5 by 2 matrix stays 5 by 2 however large or small the scalar is.
\[ 7 \cdot (5 \times 2) \text{ matrix} \;\longrightarrow\; 5 \times 2 \text{ matrix} \]
It is worth being clear about why the second option feels plausible. Multiplying is associated with getting bigger, and the entries genuinely do get bigger; but the dimensions count positions, not sizes, and no new positions appear.
This is the third operation in the lesson that leaves the shape alone, after addition and subtraction. All three work position by position, and that is exactly why none of them can change the arrangement.
Lesson 3.6 will break the pattern. Matrix multiplication builds each entry of the answer from a whole row and a whole column, and the answer's dimensions are generally different from both inputs, which is why it needs a shape rule of its own.
Socratic
The word arrives without explanation in the textbook, and it is worth a minute, because the name is describing the behaviour you have just watched.
Discussion prompt
Using the bar picture from earlier in this part, explain in two or three sentences why an ordinary number is called a scalar when it multiplies a matrix, and say what would be misleading about calling it a matrix multiplier instead.
The vocabulary earns its keep because the two kinds of multiplication really are different operations. Keeping separate names for them means that the phrase multiply the matrix by 4 can never be misread.
There is a second reason the distinction matters, which you will meet next year. A scalar can multiply any matrix at all, while a matrix can only multiply another matrix whose dimensions fit, so the two operations have completely different rules about when they exist.
If you want a one-line version to remember: a scalar rescales, a matrix rearranges. Everything in this part has been rescaling.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 188 — the definition of a scalar
Section
Part 4
Concept
Once an expression contains both a scalar and an addition, the order of operations decides the answer, and the order is the one you already use for numbers: multiplication before addition, brackets before everything.
Figure (svg): Two routes drawn as boxes. The upper route multiplies A by 4 and then adds B, giving the matrix minus 11, minus 24, 26, minus 5. The lower route adds A and B first and then multiplies by 4, giving minus 20, 0, 44, minus 20.
The two routes in the picture start from the same two matrices and end at different answers, so this is not a matter of preference. Scaling A and then adding B is what the expression written without brackets means.
\[ 4A + B \quad \text{means} \quad (4A) + B, \qquad 4(A + B) \text{ is a different matrix} \]
The book puts this in a margin note headed Compare Order of Operations, and the comparison is exactly right. If you would not read four times a plus b as four times the whole sum, do not read it that way here either.
Everything else about the two operations you already know. The addition still needs matching dimensions, the scalar multiplication still reaches every entry, and doing them in the right order is the only new demand.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 188 — the Compare Order of Operations note
Worked example
The second half of the book's Example 2 combines the two operations, and it is written out in full, one move per line, because every line is a place a mistake can enter.
\[ 4\begin{bmatrix}-2 & -8 \\ 5 & 0\end{bmatrix} + \begin{bmatrix}-3 & 8 \\ 6 & -5\end{bmatrix} \]
Read the expression before moving anything
Why: the 4 sits against the first matrix only, so it multiplies that matrix and nothing else; the second matrix is added afterwards.
\[ 4A + B \quad \text{with } A = \begin{bmatrix}-2 & -8 \\ 5 & 0\end{bmatrix}, \; B = \begin{bmatrix}-3 & 8 \\ 6 & -5\end{bmatrix} \]
Apply the scalar to all four entries of the first matrix
Why: four separate products, written out before being evaluated.
\[ \begin{bmatrix}4(-2) & 4(-8) \\ 4(5) & 4(0)\end{bmatrix} + \begin{bmatrix}-3 & 8 \\ 6 & -5\end{bmatrix} \]
Work out the scalar multiple
Why: negative eight, negative thirty-two, twenty and zero.
\[ \begin{bmatrix}-8 & -32 \\ 20 & 0\end{bmatrix} + \begin{bmatrix}-3 & 8 \\ 6 & -5\end{bmatrix} \]
Now the expression is an ordinary matrix addition of two 2 by 2 matrices, which is Example 1 again. Turning a new problem into one you have already solved is most of what algebra is.
Add the corresponding entries
Why: keep the sums unresolved for one line so that the pairing is visible.
\[ \begin{bmatrix}-8 + (-3) & -32 + 8 \\ 20 + 6 & 0 + (-5)\end{bmatrix} \]
Do the four additions
Why: negative eleven, negative twenty-four, twenty-six and negative five.
\[ \begin{bmatrix}-11 & -24 \\ 26 & -5\end{bmatrix} \]
Check by rebuilding one entry independently
Why: take the bottom left: four times five is twenty, and twenty plus six is twenty-six, which matches; and take the top right: four times negative eight is negative thirty-two, and negative thirty-two plus eight is negative twenty-four, which also matches.
\[ 4(5) + 6 = 26 \qquad 4(-8) + 8 = -24 \]
Checking two entries chosen from different rows is usually enough to catch an order-of-operations error, because that kind of mistake changes every entry at once rather than just one.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 188 — Example 2(b)
Ranking
The four moves of the example above, listed out of order. Arrange them as they must be carried out, and notice that only one arrangement gives the right answer.
Put in order
Put these four moves into the order in which they have to be done.
Why: The shape check comes first, because if the dimensions disagree there is nothing to do. Then the scalar is applied, because scalar multiplication comes before addition. Only then can corresponding entries be paired off, and the arithmetic finishes the job.
The step people are most tempted to move is the shape check, because it produces no numbers. Put it first anyway: it is the only step that can tell you the whole expression is undefined, and finding that out at the end is a waste of a page.
The last two steps look like one step, and separating them is deliberate. Writing the pairs before evaluating them means the pairing and the arithmetic can be checked independently, and each is a different kind of mistake.
Notice that this order is not special to matrices. It is the order you would use to evaluate four times a number plus another number, with one extra line at the front about shape.
Error analysis
Somebody has evaluated the same expression as Example 2(b) and got a different matrix. Every individual calculation in their working is correct, which is what makes this worth studying.
Annotate
On: \( 4A + B \;\to\; 4(A + B) = \begin{bmatrix}-20 & 0 \\ 44 & -20\end{bmatrix} \)
\[ 4\begin{bmatrix}-2 & -8 \\ 5 & 0\end{bmatrix} + \begin{bmatrix}-3 & 8 \\ 6 & -5\end{bmatrix} = \begin{bmatrix}-11 & -24 \\ 26 & -5\end{bmatrix} \]
The error is a reading error rather than a calculation error, and those are the hardest to catch by checking your arithmetic, because the arithmetic is right. Only rereading the original expression finds it.
There is a legitimate version of what they did, and it is worth knowing about. The distributive property says that four times a sum is four times the first plus four times the second, so their answer is the correct value of a different expression, one with brackets in it.
So the practical rule is to look for brackets before you start. A scalar written against one matrix touches that matrix only; a scalar written in front of a bracket touches everything inside it.
Try predicting the size of an error like this in future. Because the mistake multiplies the second matrix by 4 instead of by 1, every entry is out by three times that matrix's entry, which is a pattern you can sometimes spot in a wrong answer.
Worked example
The book's fourth checkpoint has the same shape as Example 2(b) with a different scalar, so use it to rehearse the order rather than to learn anything new.
\[ 3\begin{bmatrix}4 & -1 \\ -3 & -5\end{bmatrix} + \begin{bmatrix}-2 & -2 \\ 0 & 6\end{bmatrix} \]
Apply the scalar to the first matrix only
Why: the 3 is written against the first matrix, so the second one is untouched at this stage.
\[ \begin{bmatrix}3(4) & 3(-1) \\ 3(-3) & 3(-5)\end{bmatrix} + \begin{bmatrix}-2 & -2 \\ 0 & 6\end{bmatrix} \]
Work out the scalar multiple
Why: twelve, negative three, negative nine and negative fifteen.
\[ \begin{bmatrix}12 & -3 \\ -9 & -15\end{bmatrix} + \begin{bmatrix}-2 & -2 \\ 0 & 6\end{bmatrix} \]
Both matrices are 2 by 2, so the addition is defined and the answer will be 2 by 2 as well. It costs nothing to say that out loud at this point in the working.
Add the corresponding entries
Why: twelve plus negative two, negative three plus negative two, negative nine plus zero, and negative fifteen plus six.
\[ \begin{bmatrix}12 + (-2) & -3 + (-2) \\ -9 + 0 & -15 + 6\end{bmatrix} \]
Do the four additions
Why: ten, negative five, negative nine and negative nine.
\[ \begin{bmatrix}10 & -5 \\ -9 & -9\end{bmatrix} \]
Check the bottom row independently
Why: three times negative three is negative nine and adding zero leaves negative nine; three times negative five is negative fifteen and adding six gives negative nine, so both bottom entries are right.
\[ 3(-3) + 0 = -9 \qquad 3(-5) + 6 = -9 \]
The two bottom entries coming out equal is a coincidence of this question rather than a pattern, but it is the kind of coincidence worth noticing, because an unexpected repetition is sometimes a copied-entry error rather than a coincidence.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 188 — Guided Practice 4
Worked example
One from the exercise set with a scalar on each matrix and a subtraction between them. Take A with rows 2 and negative 1, then 0 and 4; take B with rows 1 and 3, then negative 2 and 5.
\[ 3\begin{bmatrix}2 & -1 \\ 0 & 4\end{bmatrix} - 2\begin{bmatrix}1 & 3 \\ -2 & 5\end{bmatrix} \]
Scale each matrix by its own scalar
Why: both scalar multiplications happen before the subtraction, exactly as both multiplications would happen first in an expression with numbers.
\[ \begin{bmatrix}3(2) & 3(-1) \\ 3(0) & 3(4)\end{bmatrix} - \begin{bmatrix}2(1) & 2(3) \\ 2(-2) & 2(5)\end{bmatrix} \]
Work out the two scalar multiples
Why: the first becomes six, negative three, zero and twelve; the second becomes two, six, negative four and ten.
\[ \begin{bmatrix}6 & -3 \\ 0 & 12\end{bmatrix} - \begin{bmatrix}2 & 6 \\ -4 & 10\end{bmatrix} \]
The minus sign belongs to the whole second matrix, not just to its first entry. That is why scaling it by 2 first and subtracting afterwards is safer than trying to fold the minus sign into the scalar in your head.
Subtract the corresponding entries
Why: six minus two, negative three minus six, zero minus negative four, and twelve minus ten.
\[ \begin{bmatrix}6 - 2 & -3 - 6 \\ 0 - (-4) & 12 - 10\end{bmatrix} \]
Do the four subtractions
Why: four, negative nine, four and two.
\[ \begin{bmatrix}4 & -9 \\ 4 & 2\end{bmatrix} \]
Check the entry that changed sign direction
Why: the bottom left was zero minus negative four, which is zero plus four, so four is right; and the top right was negative three minus six, which moves further down to negative nine rather than up to three.
\[ 0 - (-4) = 4 \qquad -3 - 6 = -9 \]
Those are the two entries where a sign could plausibly go wrong, so they are the two worth checking. Choosing which entries to check, rather than checking all of them, is a skill worth building on small matrices now.
Concept
Because every operation works position by position, and because each position holds an ordinary number, the familiar properties of arithmetic carry straight over to matrices of the same dimensions.
| property | what it says | why it holds |
|---|---|---|
| Associative property of addition | (A + B) + C = A + (B + C) | in each position it is just the associative property for three numbers |
| Commutative property of addition | A + B = B + A | in each position it is just a + b = b + a |
| Distributive property of addition | k(A + B) = kA + kB | in each position it is k(a + b) = ka + kb |
| Distributive property of subtraction | k(A − B) = kA − kB | in each position it is k(a − b) = ka − kb |
Read the third column and notice that no property is a new fact about matrices. Each one is an old fact about numbers, applied in every position at once, which is the general reason matrix addition behaves so well.
The distributive property is the useful one in practice. It is why the book's Another Way note can say that half of the sum of two matrices may be found either by adding first and halving after, or by halving each one and then adding.
Every property in the table needs the matrices to have the same dimensions, which the book states above the box. Without that, the sums in the statements would not exist and there would be nothing to prove.
One property is conspicuously absent, and its absence is the whole of Lesson 3.6. Multiplication of two matrices is not commutative, so A times B and B times A are usually different, and sometimes only one of them even exists.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 188 — Concept Summary: Properties of Matrix Operations
Counterexample
A property stated in a box is easy to nod at. Test this one on numbers you can check, using the DVD rack matrices from the start of the lesson, and it stops being something you have to trust.
\[ \tfrac12 (A + B) \quad \text{against} \quad \tfrac12 A + \tfrac12 B \]
Discussion prompt
Take the walnut row only: last month sold 125 small and 100 large, and this month 95 small and 114 large. Work out half of the sum of those two rows, then work out half of each row and add the results, and say what you find and why it had to happen.
This is the sort of test worth running whenever a property is handed to you. Two lines of arithmetic either confirm it or expose that you have misread it, and both outcomes are useful.
Notice which property you did not test. Nothing here shows that A times B equals B times A, and you should not assume it, because for matrices it is generally false.
The practical moral is about choosing a route. Both routes are correct, and the one that keeps the intermediate numbers tidy is the one to use, which in data problems usually means adding before scaling.
Fill the middle
One more of the same shape, worked at the level of individual entries. The expression doubles the first matrix and then subtracts the second, with the first having rows 3 and negative 4, then 6 and 0, and the second rows 5 and negative 1, then 2 and negative 7.
\[ 2\begin{bmatrix}3 & -4 \\ 6 & 0\end{bmatrix} - \begin{bmatrix}5 & -1 \\ 2 & -7\end{bmatrix} \]
Fill in the blanks
2(3) - 5 = 1, \quad 2(-4) - (-1) = -7, \quad 2(6) - 2 = 10, \quad 2(0) - (-7) = 7
Why: Each blank is one position: double the first matrix's entry, then subtract the second matrix's entry in the same position. Six minus five is 1, negative eight minus negative one is negative 7, twelve minus two is 10, and zero minus negative seven is 7. Collected up, the answer is the matrix with rows 1 and negative 7, then 10 and 7.
Two of these subtract a negative number, and both of them move the answer upward. Negative eight take away negative one is negative seven, not negative nine, and zero take away negative seven is a positive seven.
Notice that the scalar was applied before anything was subtracted, in every one of the four lines. Writing each position as a single arithmetic expression like this makes the order of operations impossible to get wrong.
Once all four boxes are right, write them back into a 2 by 2 matrix. Collecting the entries into the proper arrangement is the last step of every question in this part, and it is the step most often forgotten under time pressure.
If you would like a further check, subtract the second matrix from your answer in reverse: add the second matrix back on and you should recover the doubled first matrix, with entries six, negative eight, twelve and zero.
Section
Part 5
Concept
Everything needed to solve an equation involving matrices is already on the table: the operations from Parts 2 to 4, and the definition of equal matrices from Part 1. Putting them together is the whole method.
Figure (svg): Two two by two matrices set equal. The left has entries 15x plus 9, 15, 3 and minus 12 minus 3y; the right has minus 21, 15, 3 and minus 24. The two positions carrying letters are ringed and the equations they give are written beside them.
Because two matrices are equal only when every corresponding position agrees, one matrix equation is really several ordinary equations at once, one for each position.
Most of those positions turn out to say nothing. In the picture, two of the four positions hold only numbers, and they simply confirm that the two sides already match there. The positions carrying letters are the ones that do work.
So the plan is always the same: simplify one side until it is a single matrix, then read off an equation from each position that contains a letter, and solve those equations the ordinary way.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 190 — Solving Matrix Equations
Step zero
Before any working, decide what to do first. The expression on the left has a scalar outside a bracket containing a sum of two matrices, and the right-hand side is a single matrix.
\[ 3\left(\begin{bmatrix}5x & -2 \\ 6 & -4\end{bmatrix} + \begin{bmatrix}3 & 7 \\ -5 & -y\end{bmatrix}\right) = \begin{bmatrix}-21 & 15 \\ 3 & -24\end{bmatrix} \]
Discussion prompt
Name the first move you would make and say why, then name the move you would make after it. Say also why equating corresponding elements straight away would not work yet.
This is worth pausing on because it is the only genuinely new decision in Part 5. Once the left-hand side is one matrix, everything that follows is the equation-solving you have done since Chapter 1.
Notice that the letters change nothing about the order of the moves. An entry of 5x is added to an entry of 3 in exactly the way 5 would be, giving 5x plus 3, and the scalar then multiplies that whole entry.
Keep an eye on the entries without letters as you go. If one of those ever fails to match the right-hand side, the original equation had no solution at all, and that is worth spotting early rather than late.
Worked example
The book's final example, with every line kept visible. Two unknowns sit in different positions of the same matrix, and each will come out of its own position independently of the other.
\[ 3\left(\begin{bmatrix}5x & -2 \\ 6 & -4\end{bmatrix} + \begin{bmatrix}3 & 7 \\ -5 & -y\end{bmatrix}\right) = \begin{bmatrix}-21 & 15 \\ 3 & -24\end{bmatrix} \]
Add the two matrices inside the bracket
Why: corresponding entries, as always: 5x plus 3, negative 2 plus 7 which is 5, 6 plus negative 5 which is 1, and negative 4 plus negative y.
\[ 3\begin{bmatrix}5x + 3 & 5 \\ 1 & -4 - y\end{bmatrix} = \begin{bmatrix}-21 & 15 \\ 3 & -24\end{bmatrix} \]
The letters behave like any other entry. Adding negative y to negative 4 gives negative 4 minus y, which cannot be simplified further, so it stays as it is and waits.
Perform the scalar multiplication
Why: multiply all four entries by 3, including the ones carrying letters: three lots of 5x plus 3 is 15x plus 9, and three lots of negative 4 minus y is negative 12 minus 3y.
\[ \begin{bmatrix}15x + 9 & 15 \\ 3 & -12 - 3y\end{bmatrix} = \begin{bmatrix}-21 & 15 \\ 3 & -24\end{bmatrix} \]
Pause and compare the two matrices position by position. The entries in row 1, column 2 and row 2, column 1 already agree at 15 and 3, so those positions tell you nothing new and can be left alone.
Equate the corresponding elements that carry letters
Why: two positions remain, and each gives an ordinary linear equation in one unknown.
\[ 15x + 9 = -21 \qquad \text{and} \qquad -12 - 3y = -24 \]
Solve the equation for x
Why: subtract 9 from both sides to get 15x equal to negative 30, then divide both sides by 15.
\[ 15x = -30 \;\Longrightarrow\; x = -2 \]
Solve the equation for y
Why: add 12 to both sides to get negative 3y equal to negative 12, then divide both sides by negative 3.
\[ -3y = -12 \;\Longrightarrow\; y = 4 \]
Check by substituting both values into the original equation
Why: with x equal to negative 2 the first matrix has entries negative 10 and negative 2 on the top row, and with y equal to 4 the second matrix has negative 4 in the bottom right; adding gives negative 7 and 5 on top and 1 and negative 8 below, and multiplying all four by 3 gives negative 21, 15, 3 and negative 24, which is exactly the right-hand side.
\[ 3\begin{bmatrix}-7 & 5 \\ 1 & -8\end{bmatrix} = \begin{bmatrix}-21 & 15 \\ 3 & -24\end{bmatrix} \]
All four positions agree, not just the two that were solved, which is what makes this a genuine check rather than a restatement of the working.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 190 — Example 4
Worked example
The book's matching checkpoint, with one change that matters: the scalar outside the bracket is negative, so every sign inside gets reversed as well as every size doubled.
\[ -2\left(\begin{bmatrix}-3x & -1 \\ 4 & y\end{bmatrix} + \begin{bmatrix}9 & -4 \\ -5 & 3\end{bmatrix}\right) = \begin{bmatrix}12 & 10 \\ 2 & -18\end{bmatrix} \]
Add the two matrices inside the bracket
Why: negative 3x plus 9, negative 1 plus negative 4 which is negative 5, 4 plus negative 5 which is negative 1, and y plus 3.
\[ -2\begin{bmatrix}-3x + 9 & -5 \\ -1 & y + 3\end{bmatrix} = \begin{bmatrix}12 & 10 \\ 2 & -18\end{bmatrix} \]
Apply the scalar negative 2 to every entry
Why: negative 2 times negative 3x plus 9 is 6x minus 18, negative 2 times negative 5 is 10, negative 2 times negative 1 is 2, and negative 2 times y plus 3 is negative 2y minus 6.
\[ \begin{bmatrix}6x - 18 & 10 \\ 2 & -2y - 6\end{bmatrix} = \begin{bmatrix}12 & 10 \\ 2 & -18\end{bmatrix} \]
Distributing a negative scalar across a bracketed entry is the step to slow down for. Negative 2 multiplied by negative 3x is positive 6x, and negative 2 multiplied by 9 is negative 18, so both terms change and only one changes sign.
Equate the two positions that carry letters
Why: the other two positions already read 10 and 2 on both sides, so they confirm the working rather than extending it.
\[ 6x - 18 = 12 \qquad \text{and} \qquad -2y - 6 = -18 \]
Solve both equations
Why: add 18 to get 6x equal to 30 and divide by 6; add 6 to get negative 2y equal to negative 12 and divide by negative 2.
\[ x = 5 \qquad y = 6 \]
Check by putting both values back
Why: with x equal to 5 the top left entry inside the bracket is negative 15 plus 9, which is negative 6, and negative 2 times negative 6 is 12; with y equal to 6 the bottom right inside the bracket is 6 plus 3, which is 9, and negative 2 times 9 is negative 18.
\[ -2\begin{bmatrix}-6 & -5 \\ -1 & 9\end{bmatrix} = \begin{bmatrix}12 & 10 \\ 2 & -18\end{bmatrix} \]
Both solved positions land on the target, and so do the two that were left alone, which is the sign that the scalar was distributed correctly across every entry rather than only the ones being watched.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 190 — Guided Practice 6
Matching
Four entry equations of the kind Part 5 produces, each from a different position of some matrix equation. Match each to its solution, and notice that the matrix setting has completely disappeared by this stage.
Match the pairs
Why: Each is an ordinary two-step linear equation. Subtracting 9 and dividing by 15 gives negative 2; adding 12 and dividing by negative 3 gives 4; adding 18 and dividing by 6 gives 5; adding 6 and dividing by negative 2 gives 6. The matrix only ever supplied the equations, and solving them uses nothing from this chapter.
The two with a negative coefficient are the ones to watch. Dividing both sides by a negative number is where a sign is most often lost, and both of those equations have a negative on each side to begin with.
That is the encouraging part of Part 5. Once the matrices have been simplified and the positions equated, you are back in Chapter 1, solving equations you could already solve before this chapter started.
It also explains why the checks in the worked examples went all the way back to the original matrix equation. The entry equations can each be solved correctly while the simplification that produced them was wrong, so the only complete check is the one that rebuilds both matrices.
Section
Part 6
Concept
Return to the DVD rack company, now with the vocabulary to describe what is going on. The matrix holds the numbers and the labels sit outside it, which means every row and every column carries a meaning you supplied.
Figure (svg): The three by two sales matrix with the pine row shaded in one colour across both entries and the small column shaded in another colour down all three entries, with notes saying what each stripe means.
A row of this matrix is one kind of wood across both sizes; a column is one size across all three woods. Which is which is not written anywhere inside the brackets, and the arithmetic will happily proceed whether you remember it or not.
That is the trade matrices make. They are fast and uniform because they have forgotten what the numbers mean, so keeping the meanings straight is your job, not the notation's.
In practice that means writing the labels down beside the matrix, exactly as the book does in the Example 3 solution. Two matrices can only be sensibly added if their rows mean the same things and their columns mean the same things, and no amount of dimension checking can tell you that.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 189 — Organizing Data
Worked example
The book's multi-step problem. Last month's and this month's sales are given, and the company wants a single matrix of average monthly sales over the two-month period.
\[ A = \begin{bmatrix}125 & 100 \\ 278 & 251 \\ 225 & 270\end{bmatrix} \qquad B = \begin{bmatrix}95 & 114 \\ 316 & 215 \\ 205 & 300\end{bmatrix} \]
Organise the data as two matrices with the same labels
Why: rows walnut, pine and cherry, columns small and large, in the same order in both, so that corresponding positions really do correspond.
\[ A, B \text{ are both } 3 \times 2 \]
That first step is the one that carries all the risk. Once the two matrices are laid out with the same rows and the same columns, everything after it is arithmetic that cannot go conceptually wrong.
Add the two matrices to get the two-month total
Why: corresponding entries: 125 plus 95 is 220, 100 plus 114 is 214, 278 plus 316 is 594, 251 plus 215 is 466, 225 plus 205 is 430, and 270 plus 300 is 570.
\[ A + B = \begin{bmatrix}220 & 214 \\ 594 & 466 \\ 430 & 570\end{bmatrix} \]
Multiply the total by one half to get the average
Why: an average of two numbers is their total halved, and scalar multiplication applies that to all six positions at once.
\[ \tfrac12 (A + B) = \begin{bmatrix}110 & 107 \\ 297 & 233 \\ 215 & 285\end{bmatrix} \]
Read the answer back into words before finishing. The company sold an average of 110 small walnut racks, 107 large walnut, 297 small pine, 233 large pine, 215 small cherry and 285 large cherry per month.
Check one entry against the definition of an average
Why: take the small pine entry: the two months were 278 and 316, whose total is 594, and half of 594 is 297, which is what the matrix says; and each average sits between the two monthly figures, as an average must.
\[ \tfrac{278 + 316}{2} = \tfrac{594}{2} = 297 \]
The between test is worth running on every entry at a glance. An average that falls outside the two numbers it came from is impossible, so it catches any entry where an addition or a halving went wrong.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 189 — Example 3
Picture it
Before subtracting anything, look at the small-rack figures as a picture. The three woods did not move in the same direction, and that is precisely the information a single total would destroy.
Figure (svg): A grouped bar chart of small rack sales for walnut, pine and cherry, with last month's bars at 125, 278 and 225 and this month's at 95, 316 and 205.
Pine rose from 278 to 316, while walnut fell from 125 to 95 and cherry fell from 225 to 205. Added together the three woods barely moved, and the company would learn nothing from that one number.
A matrix keeps all six figures separate, so it can record all six changes separately. That is the real argument for organising data this way rather than summarising it early.
It also tells you what to expect from the difference of the two matrices on the next slide: one positive entry among the small racks and two negative ones, with sizes of roughly 30, 38 and 20.
Worked example
The checkpoint asks for this month minus last month, and then for an interpretation. Both parts matter, and the second is the one that makes the first worth doing.
\[ B - A = \begin{bmatrix}95 & 114 \\ 316 & 215 \\ 205 & 300\end{bmatrix} - \begin{bmatrix}125 & 100 \\ 278 & 251 \\ 225 & 270\end{bmatrix} \]
Check the shapes and set out the six differences
Why: both are 3 by 2 with the same labels in the same order, so the difference is defined and every position compares like with like.
\[ \begin{bmatrix}95 - 125 & 114 - 100 \\ 316 - 278 & 215 - 251 \\ 205 - 225 & 300 - 270\end{bmatrix} \]
Work out the six entries
Why: negative 30 and 14, then 38 and negative 36, then negative 20 and 30.
\[ B - A = \begin{bmatrix}-30 & 14 \\ 38 & -36 \\ -20 & 30\end{bmatrix} \]
Now read it. A positive entry is a rise from last month to this month and a negative entry is a fall, so this matrix records the change in sales for each wood and size separately.
Walnut lost 30 small racks and gained 14 large ones; pine gained 38 small and lost 36 large; cherry lost 20 small and gained 30 large. In every wood the two sizes moved in opposite directions, which is a pattern worth a sales manager's attention.
Check by adding the difference back on to last month
Why: 125 plus negative 30 is 95, 100 plus 14 is 114, 278 plus 38 is 316, 251 plus negative 36 is 215, 225 plus negative 20 is 205, and 270 plus 30 is 300, which rebuilds this month's matrix exactly.
\[ A + (B - A) = \begin{bmatrix}95 & 114 \\ 316 & 215 \\ 205 & 300\end{bmatrix} \]
Had the subtraction been taken the other way round, every entry would have the opposite sign and the matrix would describe the change from this month back to last month, which is a different and equally answerable question.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 190 — Guided Practice 5
Picture it
Example 3 did two operations to six numbers at once. Follow a single entry through both of them and the notation stops looking clever and starts looking like bookkeeping, which is all it is.
Figure (svg): A flow diagram tracing the walnut small-rack entry: 125 from last month and 95 from this month feed a plus node, giving a total of 220, which then passes through a halving node to give an average of 110.
The walnut small-rack figure went in as 125 and 95, was added to 220, and was halved to 110. Every other entry took exactly the same journey with different numbers, and none of them interacted with any other.
Written out in ordinary arithmetic that would be six separate add-and-halve calculations, each needing its own line and its own label. The matrix version is those six calculations stacked, with the labels written once.
That saving is the entire practical argument for matrices at this stage of the course. With six entries it is convenient; with a company's twelve months of sales across forty products it is the difference between possible and not.
Real world
Stretch the same idea until it stops being a textbook exercise. Suppose the company keeps one 3 by 2 matrix like these for every month of the year.
Discussion prompt
Say what adding all twelve of those matrices would give, what multiplying that total by one twelfth would give, and name one question about the year's sales that neither matrix addition nor scalar multiplication could answer.
The middle answer is the one worth dwelling on. Every operation in this lesson is a way of combining tables that have the same shape and the same meanings, and combining always throws away the distinctions between the things combined.
That is not a flaw; it is the point. You add the twelve months precisely because you want the year rather than the months, and you keep the twelve matrices as well precisely because you will sometimes want the months back.
The third answer points straight at the next lesson. The moment you want to multiply counts by prices, you need an operation that mixes a whole row with a whole column, and none of the three operations in this lesson does that.
McDougal Littell Algebra 2 (Texas Edition), Ch. 3 Linear Systems and Matrices — Lesson 3.5 Perform Basic Matrix Operations §3.5, p. 189 — the manufacturing context of Example 3
Socratic
A matrix is a very efficient object, and efficiency is always paid for. Say exactly what the payment is here, because knowing it is what stops matrices being misused.
Discussion prompt
Compare the original sales table with the matrix that came from it. Name two things the table told you that the matrix does not, and explain why matrix addition can produce a perfectly valid matrix that means nothing at all.
This is the single most important idea in Part 6, and it is not really about matrices at all. Any time data is stripped down for calculation, something has to keep track of what the numbers meant.
It also explains a habit you will see in every worked example in the book. The labels are written beside the matrix in Example 3 and referred to again in the interpretation step, so the meaning is carried alongside the arithmetic rather than being trusted to memory.
Carry the habit into your own work. Write the row and column labels next to any matrix you build from data, and read the answer back into words at the end. Both take seconds and both catch the kind of error no dimension check can.
Section
Part 7
Pattern
Everything in this lesson fits into one short routine. The order matters, and the first step is the one that is skipped most often and costs the most when it is.
The routine is short because the operations are genuinely simple. Everything in this lesson is arithmetic you learned years ago, applied to several positions at once with the bookkeeping written down.
If you only remember one line of it, remember the first. Shape decides whether an expression means anything at all, and it decides it before any arithmetic is attempted.
Comparison
Most of what you know about arithmetic transfers, and the few places it does not are worth naming. Fill the gaps from what you have seen in this lesson.
Comparison matrix
| question | with ordinary numbers | with matrices |
|---|---|---|
| when can two be added? | always | same dimensions |
| does order matter in a sum? | no | no |
| does multiplying by 4 change the size of the object? | there is no shape to change | no |
| what makes two of them equal? | the same value | same dimensions and every position agreeing |
| what comes first, scaling or adding? | scaling | scaling |
Read down the last column and notice how little is genuinely new. Three of the five rows behave exactly as numbers do, because every operation in this lesson works one position at a time on ordinary numbers.
The row that differs is the first, and it differs because matrices have a shape and numbers do not. Everything awkward in Part 2 traces back to that single fact.
The equality row is worth reading twice as well. Two numbers are equal when they have the same value, but a matrix has no single value, so equality has to be defined position by position instead.
Keep this table in mind when Lesson 3.6 arrives, because that is where the second column and the third genuinely part company: with numbers the order of a product never matters, and with matrices it almost always does.
Check
Decide on the shapes alone. Remember that both numbers of the dimensions have to agree, and that the number of elements is not the test.
Check your understanding
Which of these additions is defined?
Answer: C
Why: Addition pairs off positions, so it needs both dimensions to agree. Only the pair with three rows and four columns on each side has a partner for every position, and its sum is another 3 by 4 matrix.
Three of the four options hold the same number of elements on each side, which is exactly why counting elements is not a test. Two of them even come from the same pair of matrices written the other way round.
The quickest reliable method is to say both pairs of dimensions out loud and compare them as pairs rather than as four separate numbers. A 3 by 4 and a 3 by 4 match; a 3 by 4 and a 4 by 3 do not.
When the answer is that a sum is undefined, write that down as your answer rather than leaving the question blank. It is a complete response and it is what the book's Guided Practice asks for when it says to perform the operation if possible.
One more habit worth building from this question: once you know a sum is defined, you also know the shape of the answer, because it is the shape the two inputs share.
Check
The book's Example 1(b) again, reduced to a single position. The first matrix has negative 2 in row 2, column 2, and the second has negative 10 there.
Check your understanding
A subtraction has negative 2 in row 2, column 2 of the first matrix and negative 10 in the same position of the second. What is that entry of the difference?
Answer: A
Why: Subtracting a negative number is adding its opposite, so negative 2 minus negative 10 is negative 2 plus 10, which is 8. Every entry of a matrix difference is worked out exactly this way, using only the two numbers in that one position.
Every wrong option here is a real mistake with a name. One forgets that subtracting a negative adds, one reverses the order of the subtraction, and one loses a sign on the way.
A number line settles it if you are unsure. Starting at negative 2 and removing a debt of 10 moves you up to positive 8, whereas adding a debt of 10 would move you down to negative 12.
Notice that nothing about the rest of the matrix was needed. Every position of a matrix difference is computed in isolation, which is why a question like this can be asked about a single entry at all.
Order matters here in a way it did not for addition. The difference of two matrices taken the other way round has every entry negated, so read the question carefully to see which one is being subtracted from which.
Check
An expression with a scalar on each matrix. Take A with rows 2 and negative 1, then 0 and 4, and take B with rows 1 and 3, then negative 2 and 5, and look only at the entry in row 1, column 2.
Check your understanding
For the expression three times A minus two times B, what is the entry in row 1, column 2?
Answer: A
Why: That position holds negative 1 in A and 3 in B. Scaling first gives three times negative 1, which is negative 3, and two times 3, which is 6, and then subtracting gives negative 3 minus 6, which is negative 9.
The last wrong option is the order-of-operations error from Part 4 in miniature. Subtracting before scaling gives a number that looks reasonable and is wrong by a wide margin.
Working one position at a time like this is a good way to handle any mixed expression under time pressure. Each entry of the answer is a single line of ordinary arithmetic, and no entry depends on any other.
If you want the whole answer, the same routine in the other three positions gives 4, then 4 and 2, so the matrix has rows 4 and negative 9, then 4 and 2.
The check is the same one used in the worked example. Rebuild one entry from scratch, starting from the original expression rather than from your own intermediate lines, and see whether it lands on the same number.
Explain it to yourself
One sentence has run through this whole lesson, and putting it in your own words is what makes it portable. Write your answer before reading the one below.
Discussion prompt
In three or four sentences, explain why adding matrices is easier than adding fractions but fussier about when it is allowed, and use the words positions and dimensions. Then say what that tells you to check first in any matrix question.
If your explanation talked mostly about the arithmetic, write it again with the word positions doing more work. The arithmetic is the easy part and is not really what this lesson was about.
A good test of whether you have it: could you explain to somebody who has never seen a matrix why a 2 by 3 and a 3 by 2 cannot be added, without writing down a single number? If yes, you have the idea.
The same sentence will carry you into Lesson 3.6 as well, where a different pairing rule decides when a product exists. The habit of asking about shape before arithmetic is the transferable part of today.
Exit ticket
One question using every part of the lesson: a shape check, a scalar, a subtraction, and a reading of the answer in the order the positions are named.
\[ 4\begin{bmatrix}-2 & -8 \\ 5 & 0\end{bmatrix} - 2\begin{bmatrix}1 & -3 \\ 4 & -6\end{bmatrix} \]
Discussion prompt
State the dimensions of the answer before working anything out, then evaluate the expression, and finally name the entry in row 2, column 1 and say which two numbers from the question produced it.
Do the shape check genuinely rather than as a formality. Saying two by two out loud before starting is what turns it into a habit that will save you when the shapes do not match.
Both scalars must be applied before the subtraction, and the second scalar multiplies the whole of the second matrix, not just its first entry. Those are the two places this question can go wrong.
The bottom right entry is worth a second look on its own. Zero minus negative twelve is positive twelve, which is the subtracting-a-negative rule one last time.
Finish by reading the answer back position by position, the way Example 3 reads its answer back into racks. It is the last chance to notice an entry that has been dropped or written in the wrong place.
Recap
A matrix is a table of numbers with the labels moved outside, described by its dimensions and addressed by row and then column. Three operations act on it, all of them working one position at a time.
| operation | what it does | when it is defined | shape of the answer |
|---|---|---|---|
| addition | adds the entries in corresponding positions | only when the dimensions match exactly | the same as both inputs |
| subtraction | subtracts the entries in corresponding positions | only when the dimensions match exactly | the same as both inputs |
| scalar multiplication | multiplies every entry by one number | always | unchanged |
| equating two matrices | gives one equation per position | only when the dimensions match exactly | no matrix; a set of ordinary equations |
\[ \begin{bmatrix}a & b \\ c & d\end{bmatrix} + \begin{bmatrix}e & f \\ g & h\end{bmatrix} = \begin{bmatrix}a+e & b+f \\ c+g & d+h\end{bmatrix} \]
The line above is the shape to carry away. Every position of the answer is built from the two entries in that same position, and once you believe that, the rules about dimensions stop being arbitrary and start being obvious.
Read the last row of the table again as well. Setting two matrices equal is not a fourth operation so much as a way of turning one matrix statement into several ordinary equations, which is what Example 4 exploited.
Lesson 3.6 changes the picture. Multiplying two matrices builds each entry of the answer from a whole row and a whole column, so the dimension rule becomes different, the answer's shape becomes different, and the order of the two matrices starts to matter.
Everything from today survives into that lesson and into Lessons 3.7 and 3.8 on determinants and inverses. Dimensions written rows first, elements addressed row then column, and the habit of checking the shape before starting are all assumed from here on.
OpenStax Algebra and Trigonometry 2e, §11.5 Matrices and Matrix Operations §11.5 — the same three operations, with the entry notation written as a with a row and column subscript
Want this taught 1-on-1? Alexander tutors Algebra 2 — $55/session, free consultation.