1
$\begingroup$

I have the following matrix: $$ A = \begin{bmatrix} 2 & 1 & 1 & 0 \\ 4 & 3 & 2 & 1 \\ 8 & 7 & 8 & 5 \\ 6 & 7 & 9 & 8 \\ \end{bmatrix} $$

and I want to apply Gauss elemination through a series of elementary matrix multiplications.

To cancel out the numbers under the first diagonal element of $A$, $$E_1 = \begin{bmatrix} 1 & 0 & 0 & 0 \\ -2 & 1 & 0 & 0 \\ -4 & 0 & 1 & 0 \\ -3 & 0 & 0 & 1 \\ \end{bmatrix} $$

Applying $E_1$, it works as expected

$$A^1 = E_1A = \begin{bmatrix} 2 & 1 & 1 & 0 \\ 0 & 1 & 0 & 1 \\ 0 & 3 & 4 & 5 \\ 0 & 4 & 6 & 8 \\ \end{bmatrix}$$

For the second column, however, starting with $E_2 = \begin{bmatrix} 1 & 0 & 0 & 0 \\ 0 & 1 & 0 & 0 \\ 0 & 0 & 1 & 0 \\ 0 & 0 & 0 & 1 \\ \end{bmatrix}$, how do I eliminate numbers under the second diagonal element of $A$?

If I want to apply on $E_2$ the row operations that would make the numbers under the second diagonal element of $A$ zero, then I would get for the third row of $E_2$, assuming it starts as an identity matrix, $I_n$ wouldn't I get:

$$E_2 = \begin{bmatrix} 1 & 0 & 0 & 0 \\ 0 & 1 & 0 & 0 \\ -7 & 0 & 1 & 0 \\ -7 & 0 & 0 & 1 \\ \end{bmatrix}$$

The row operations in the context of A are apply $-7R_1 + R_3$ and $-7R_1 + R_4$ to cancel out the numbers under the second diagonal element of $A$ - so, the above shows I applied those to $I_n$

Doing the multiplication though, I get

$$A^2 = E_2E_1A = \begin{bmatrix} 2 & 1 & 1 & 0 \\ 0 & 1 & 0 & 1 \\ -14 & -4 & -3 & 5 \\ -14 & -3 & -2 & 8 \\ \end{bmatrix}$$

What am I doing wrong?

Edit: Calculation error in $A^1$ at position $(4,3)$. It should be $9 + (-3)\times 1 = 6 \neq 5$

$\endgroup$

1 Answer 1

2
$\begingroup$

First of all, you are applying $E_2$ to $A^1$. Hence, while designing $E_2$, you should be looking at $A^1$ (and not at $A$). While looking at $A_1$, you know that you want to cancel $3$ and $4$ at position $(3,2)$ and $(4,2)$ respectively (and not $-7$ of $A$).

You can try $E_2$ as follows: $$E_2 = \left[ \begin{array}{cccc} 1& 0 & 0&0\\ 0&1&0&0\\ 0&-3&1&0\\ 0&-4&0&1 \end{array} \right] $$

The reason to select this $E_2$ is that you want to operate $-3R_2 + R_3$ at $R_3$ and $-4R_2 + R_4$ at $R_4$. Hence, the third row of $E_2$ should have $-3$ and $+1$ in the second and third column respectively. Similarly, the fourth row of $E_2$ should have $-4$ and $+1$ in the second and fourth column.

In addition to the reason mentioned in the beginning, your $E_2$ also doesn't work because $-7$ in your $E_2$ operates on first column. If you want to limit the operation on the second column, the first column of $E_2$ should remain $[1\ 0\ 0\ 0]$.

At the end, you should get $$A^2 = E_2 E_1 A = \left[ \begin{array}{cccc} 2& 1 & 1&0\\ 0&1&0&1\\ 0&0&4&2\\ 0&0&6&4 \end{array} \right] $$

Now, you can design $E_3$ by using the operation $-\frac{3}{2}R_3 + R_4$ at $R_4$ of $A^2$.

$\endgroup$
2
  • $\begingroup$ If I was doing this without elementary matrices and was applying direct row operations on $A$ , wouldn't $-7R_1 + R_3$ and $-7R_1 + R_4$ work? That is, if I cleared out the sub diagonal elements in column 1? $\endgroup$
    – Art
    Feb 9, 2019 at 3:04
  • 1
    $\begingroup$ So, it is important to check if you are operating on $A$, or $A^1$. Also, once you have cleared the first column of $A$, you should be avoiding to use $R_1$ operations. All $R_1$ operations will add some numbers in the first column (like $-7$) and if you try to clear them, you will get the original matrix. You can try for your understanding. $\endgroup$
    – KK Patel
    Feb 9, 2019 at 3:08

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.