Miscellaneous · Q3: Find with
Find the values of , , if the matrix
satisfy the equation A′A = I.
Let . Then
As , equating corresponding entries,
, , .
Want the reason behind every step, the basics and the checks?
Row of is column of A. So the entry of is column of A times column of A, multiplied entry by entry and added. Here the three columns of A are arranged so that every product of two different columns cancels to 0. Only the diagonal survives, giving as a diagonal matrix with , , on the diagonal, and setting it equal to I gives three simple equations.
- Given
- The matrix A above, with real unknowns , and the condition (the identity).
- To find
- All real for which .
- Plan
- (1) Write down and check the orders. (2) Compute all nine entries of . (3) Equate entry by entry with I. (4) Solve each equation, keeping both signs. (5) Check the answers.
Basics you need
The rows of are the columns of A, in order. A matrix has a transpose. Basics: reading a table sideways
The entry of is . Basics: pair up, multiply, add
Two matrices of the same order are equal exactly when every pair of corresponding entries is equal. Lesson: equality of matrices
For , has the two real solutions . Forgetting the minus sign loses half the answers. Basics: solving equations
Solution
Write down and check the orders
The columns of A are
and these become the rows of the transpose:
is and A is : columns of rows of A, so is defined and is , the same order as . The equation therefore makes sense.
We read the transpose off column by column rather than trying to "flip" the picture in one go, because that avoids sign slips with the entries and .Row 1 of
Row 1 of is . Multiply it by each column of A:
In the terms and cancel; in the terms and cancel. These cancellations happen for every value of , not just special ones, because of how the signs in A are placed.Row 2 of
Row 2 of is :
In all three terms are squares, so they cannot cancel: . In the first term is cancelled by the two terms . Also , as it must be, because is always symmetric: .: the product of two negatives is positive. Basics: pair up, multiply, addRow 3 of , and the whole product
Row 3 of is :
Assembling the nine entries:
The product is a diagonal matrix for every choice of . The diagonal entries are the "lengths squared" of the columns (each column times itself), and the zeros say that different columns are "perpendicular" (their products cancel). This is exactly why the condition is easy to meet here.A diagonal matrix has all its non-diagonal entries zero (Section 3.3 (iv)). Lesson: types of matricesEquate with I, entry by entry
Both matrices are , so they are equal exactly when all nine pairs of entries agree. The six off-diagonal pairs read , which is true whatever are, so they impose no condition. The three diagonal pairs give
It is worth saying out loud that the off-diagonal equations are automatically satisfied. If one of them had come out as, say, , it would have been an extra condition and could have ruled out some of the solutions below.Definition 2 (Section 3.3.1): equal matrices have equal orders and equal corresponding entries. Lesson: equality of matricesSolve each equation, keeping both signs
The three equations are independent of one another (each involves one unknown), so the signs can be chosen freely: there are solutions in all. In rationalised form, , , .
A positive number has two real square roots. Every sign choice works because only , , appear in . Changing the sign of changes the sign of the first column of A, and that does not change any of the column products in Steps 2 to 4.
exactly when
with the three signs chosen independently (8 solutions). (Textbook answer: , , .)
Take all plus signs: , , . Then
✓. With exact values, , , give . The verification script substitutes all 8 sign choices into A and confirms exactly for each.
- Computing instead of . The order matters. (Here is not diagonal in general.)
- Giving only the positive roots. has two solutions.
- Sign slips in , especially the and entries. Read the columns of A one at a time.
- Ignoring the off-diagonal equations without checking them. They happen to be here, but you only know that after computing them.
- The entry of is column of A times column of A. This turns the product into a set of "column products", and is always symmetric.
- means every column times itself gives 1 and different columns give 0. Such a matrix is called orthogonal (a name you will meet later).
- When an equation involves only squares, both signs of each root are solutions, and independent choices multiply the number of solutions.
Every fact used, and where it comes from
| Fact | Used in | Source |
|---|---|---|
| Transpose: | Step 1 | Definition 3, Section 3.5 (Lesson: transpose) |
| Product entry formula; order of a product | Steps 1–4 | Section 3.4.5 (Lesson: matrix multiplication) |
| is symmetric | Step 3 | Section 3.5.1, properties (i) and (iv) |
| Diagonal and identity matrices | Steps 4, 5 | Section 3.3 (iv) and (vi) (Lesson: types of matrices) |
| Equality of matrices | Step 5 | Definition 2, Section 3.3.1 (Lesson: equality of matrices) |
| All eight solutions; decimals | Step 6, Check | Solved and substituted with sympy in the site's verification script |
Tell us if something on this page is wrong or unclear, or if you would like a fuller explanation, another method or a new feature. Send feedback on this problem →