Find the basis so that the matrix will be diagonal

In summary, we can show that the matrix ##TD(1) = 0##, the matrix ##TD(e_1) = x##, the matrix ##TD(e_2) = x^2##, the matrix ##TD(e_3) =0##, and the matrix ##TD(e_4) =0##.
  • #1
Hall
351
88
Homework Statement
Let ##V## be the Linear space of all real polynomials of degree ##\leq3##. Let D denote the differentiation operator and let ##T : V \to V## be the Linear transformation which maps ##p(x)## onto ##x p'(x)##.

Let ##W## be the image of ##V## under ## TD##. Find the bases for V and W relative to which the matrix of ##TD## is in diagonal form.
Relevant Equations
See the main body, please
First of all, it is clear that we can find such a bases (the theorem is given in almost all of the books, but if you want to share some insight I shall be highly grateful.)

We can show that ##W## will be the set of all real polynomials with degree ##\leq 2##. So, let's have ##\{1,x,x^2\}## as the basis for ##W##.

We want this matrix (having this idea that V is four dimensional):
$$
\begin{bmatrix}
1 &0&0&0\\
0&1&0&0\\
0&0&1&0
\end{bmatrix}
$$
Let the basis for ##V## be ##\{e_1,e_2,e_3,e_4\}##. Then, ##e_1 = c_1 + c_2x +c_3x^2 + c_4x^3##, we have
$$
TD (e_1)= 1 $$
Which implies
$$
6c_4 x^2 + 2c_3x -1=0$$

But I cannot solve c's. Now, I seek for your gentle guidance.
 
Last edited:
Physics news on Phys.org
  • #2
Hall said:
Homework Statement:: Let ##V## be the Linear space of all real polynomials of degree ##\leq3##. Let D denote the differentiation operator and let ##T : V \to V## be the Linear transformation which maps ##p(x)## onto ##x p'(x)##.

Let ##W## be the image of ##V## under ## TD##. Find the bases for V and W relative to which the matrix of ##TD## is in diagonal form.
Relevant Equations:: See the main body, please

We can show that ##W## will be the set of all real polynomials with degree ##\leq 2##. So, let's have ##\{1,x,x^2\}## as the basis for ##W##.
What polynomial maps to the polynomial ##1##?
Hall said:
We want this matrix (having this idea that V is four dimensional):
$$
\begin{bmatrix}
1 &0&0&0\\
0&1&0&0\\
0&0&1&0
\end{bmatrix}
$$
I would say that only a square matric can be diagonal.
Hall said:
Let the basis for ##V## be ##\{e_1,e_2,e_3,e_4\}##. Then, ##e_1 = c_1 + c_2x +c_3x^2 + c_4x^3##, we have
$$
TD (e_1)= 1 $$
Which implies
$$
6c_4 x^2 + 2c_3x -1=0$$

But I cannot solve c's. Now, I seek for your gentle guidance.
Perhaps rethink this approach.

It's a strange question. Are you sure it's not supposed to be ##DT##?
 
  • #3
PeroK said:
What polynomial maps to the polynomial 1
Well, I don't think (I mean I cannot find) there is a polynomial ##p(x)## such that ## x p''(x)=1##.

PeroK said:
I would say that only a square matric can be diagonal.
Sometimes, I have heard the book to say "the main diagonal".

PeroK said:
a strange question. Are you sure it's not supposed to be DT?
Yes, it is TD.
Screenshot_20220212-220721_Adobe Acrobat.jpg


Question number 20. (Apostle's Calculus Vol. 2)
 
  • #4
If you stick with the basis ##1, x, x^2, x^3## for ##V##, can you find a suitable basis for ##W##?
 
  • #5
PeroK said:
If you stick with the basis ##1, x, x^2, x^3## for ##V##, can you find a suitable basis for ##W##?
##TD(1)=0##, and that may make the whole first column zero, as zero element cannot fall in the basis for W.
 
  • #6
Hall said:
##TD(1)=0##, and that may make the whole first column zero, as zero element cannot fall in the basis for W.
Well, okay. The diagonal entries don't have to be non-zero.
 
  • #7
Hall said:
##TD(1)=0##, and that may make the whole first column zero, as zero element cannot fall in the basis for W.

This is going to be unavoidable, since the domain is higher dimension than the codomain
 
  • #8
So, should we consider "the main diagonal" as starting from the first element of the second column?

Should we seek for
$$
\begin{bmatrix}
0 &1& 0& 0\\
0 &0 &1&0\\
0 & 0 &0 &1
\end{bmatrix}
$$
?
 
  • #9
Hall said:
So, should we consider "the main diagonal" as starting from the first element of the second column?

Should we seek for
$$
begin{bmatrix}
0 &1& 0& 0\\
0 &0 &1&0\\
0 & 0 &0 &1
\end{bmatrix}
$$
?
I don't think the question makes much sense unless we consider ##TD: V \to V##.

Asking for the dimension of the range of ##TD## would be a better question to get you started.
 
  • #10
PeroK said:
I don't think the question makes much sense unless we consider ##TD: V \to V##.

Asking for the dimension of the range of ##TD## would be a better question to get you started.
Yes, of course the dimension of W is 3 and that of V is 4. But that's a very nice point to consider TD as a mapping from V to V, not to TD(V).

But still, if we consider the basis for the codomain V as ##\{1, x, x^2, x^3\}##, we would still get the issue of finding a polynomial such that ##x p''(x)=1##.

I'm getting this feeling that TD(V) shall not contain 1.
 
  • #11
Hall said:
Yes, of course the dimension of W is 3
Is it?
Hall said:
But still, if we consider the basis for the codomain V as ##\{1, x, x^2, x^3\}##, we would still get the issue of finding a polynomial such that ##x p''(x)=1##.
Why consider the codomain first? I'd look first at the action of ##TD## on the usual basis for ##V##.
Hall said:
I'm getting this feeling that TD(V) shall not contain 1.
It's more than a feeling!
 
  • #12
How about the following one?

Basis for W ##\{x,x^2\}##. Let the basis for ##V## be ##\{e_1, e_2, e_3, e_4\}##. And we want the matrix to look like
$$
\begin{bmatrix}
1 &0&0&0\\
0&1&0&0
\end{bmatrix}
$$
##TD(e_1) = x##, by inspection I can say ##e_1= x^2/2##.

##TD(e_2)= x^2## by inspection we have ##e_2= x^3/6##

## TD (e_3) =0##, we can easily get ##e_3 =1##

## TD(e_4)=0##, I think we have ## e_4=x##
 
  • Like
Likes PeroK
  • #13
I think that noting [tex]
TD : x^k \mapsto k(k-1)x^{k-1}[/tex] gets you there immediately: [itex]TD(\{1,x,x^2,x^3\}) = \{0, 2x, 6x^2\}[/itex] must span [itex]W[/itex], so a basis for [itex]W[/itex] is [itex]\{TD(x^2) = 2x, TD(x^3) = 6x^2\}[/itex].
 
  • Like
Likes Hall
  • #14
pasmith said:
I think that noting [tex]
TD : x^k \mapsto k(k-1)x^{k-1}[/tex] gets you there immediately: [itex]TD(\{1,x,x^2,x^3\}) = \{0, 2x, 6x^2\}[/itex] must span [itex]W[/itex], so a basis for [itex]W[/itex] is [itex]\{TD(x^2) = 2x, TD(x^3) = 6x^2\}[/itex].
I just wanted to hear from you how did you ensure
[tex]
TD (\{1, x, x^2, x^3\}) = \{0, 2x, 6x^2\}
[/tex]
span W? I mean do we have theorem or a something which says that Linear transformation when applied to the basis of its domain results in values which we would span the range?
 
  • #15
Hall said:
I mean do we have theorem or a something which says that Linear transformation when applied to the basis of its domain results in values which we would span the range?
It follows from the linearity of ##T##:
$$T(v) = T(v_1e_1 + v_2e_2 \dots + v_ne_n) = v_1T(e_1) + \dots + v_nT(e_n)$$
 
  • Like
Likes Hall and pasmith

FAQ: Find the basis so that the matrix will be diagonal

1. What is a basis in linear algebra?

A basis in linear algebra is a set of vectors that can be used to represent any other vector in a vector space through linear combinations. It is a fundamental concept in linear algebra and is used to understand and solve problems related to vector spaces, matrices, and transformations.

2. How do you find the basis for a matrix?

To find the basis for a matrix, you need to first reduce the matrix to its reduced row echelon form (RREF). The columns of the RREF matrix that contain the leading 1's form the basis for the matrix. These columns represent the pivot columns of the original matrix and are linearly independent, meaning they can be used to represent any other column in the matrix.

3. Why is it important to find the basis for a matrix?

Finding the basis for a matrix allows us to understand the properties and behavior of the matrix. It also helps us to solve problems related to linear transformations and systems of linear equations. In addition, the basis can be used to find the rank and nullity of a matrix, which are important concepts in linear algebra.

4. Can a matrix have more than one basis?

Yes, a matrix can have multiple bases. This is because there can be different sets of linearly independent vectors that can be used to represent all the other vectors in the matrix. However, the number of vectors in each basis will be the same, which is equal to the number of columns (or rows) in the matrix.

5. How do you diagonalize a matrix using its basis?

To diagonalize a matrix using its basis, we need to first find the basis for the matrix. Then, we can construct a transformation matrix using the basis vectors as its columns. This transformation matrix can be used to convert the original matrix into its diagonal form, where all the off-diagonal elements are zero. This process is known as a change of basis and is an important technique in solving problems related to diagonal matrices.

Similar threads

Back
Top