Scribd
Upload
127 views1 page

Linear Algebra and Its Application.130 PDF

1. This section discusses the four fundamental vector spaces associated with a matrix A: the column space, nullspace, row space, and left nullspace. 2. The dimensions of these four spaces are related by the rank r of the matrix A, such that the dimension of the column/row spaces is r, and the dimensions of the nullspaces are the number of rows/columns minus r. 3. An example is provided to illustrate a 2x2 matrix A and its associated vector spaces.

Uploaded by

raqibapp
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
127 views1 page

Linear Algebra and Its Application.130 PDF

1. This section discusses the four fundamental vector spaces associated with a matrix A: the column space, nullspace, row space, and left nullspace. 2. The dimensions of these four spaces are related by the rank r of the matrix A, such that the dimension of the column/row spaces is r, and the dimensions of the nullspaces are the number of rows/columns minus r. 3. An example is provided to illustrate a 2x2 matrix A and its associated vector spaces.

Uploaded by

raqibapp
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 1

120 Chapter 2 Vector Spaces

This law applies equally to AT , which has m columns. AT is just as good a matrix as A.
But the dimension of its column space is also r, so

r + dimension N (AT ) = m. (1)

2P The left nullspace N (AT ) has dimension m r.

The m r solutions to yT A = 0 are hiding somewhere in elimination. The rows of A


combine to produce the m r zero rows of U. Start from PA = LU, or L1 PA = U. The
last m r rows of the invertible matrix L1 P must be a basis of ys in the left nullspace
because they multiply A to give the zero rows in U.
In our 3 by 4 example, the zero row was row 3 2(row 2) + 5(row 1). Therefore
the components of y are 5, 2, 1. This is the same combination as in b3 2b2 + 5b1 on
the right-hand side, leading to 0 = 0 as the final equation. That vector y is a basis for
the left nullspace, which has dimension m r = 3 2 = 1. It is the last row of L1 P,
and produces the zero row in Uand we can often see it without computing L1 . When
desperate, it is always possible just to solve AT y = 0.
I realize that so far in this book we have given no reason to care about N (AT ). It is
correct but not convincing if I write in italics that the left nullspace is also important. The
next section does better by finding a physical meaning for y from Kirchhoffs Current
Law.
Now we know the dimensions of the four spaces. We can summarize them in a table,
and it even seems fair to advertise them as the

Fundamental Theorem of Linear Algebra, Part I


1. C (A) = column space of A; dimension r.
2. N (A) = nullspace of A; dimension n r.
3. C (AT ) = row space of A; dimension r.
4. N (AT ) = left nullspace of A; dimension m r.
" #
1 2
Example 1. A = has m = n = 2, and rank r = 1.
3 6

1. The column space contains all multiples of 13 . The second column is in the same
direction and contributes nothing new.

2. The nullspace contains all multiples of 2 1 . This vector satisfies Ax = 0.

3. The row space contains all multiples of 12 . I write it as a column vector, since
strictly speaking it is in the column space of AT .

4. The left nullspace contains all multiples of y = 3 1 . The rows of A with coeffi-
T
cients 3 and 1 add to zero, so A y = 0.

You might also like