Jordan normal form

The JordanChevalley decomposition is particularly simple with respect to a basis for which the operator
takes its Jordan normal form. The diagonal form for
diagonalizable matrices, for instance normal matrices, is
a special case of the Jordan normal form.[5][6][7]
The Jordan normal form is named after Camille Jordan.

1 Overview
1.1 Notation
Some textbooks have the ones on the subdiagonal, i.e.,
immediately below the main diagonal instead of on the
superdiagonal. The eigenvalues are still on the main
diagonal.[8][9]

An example of a matrix in Jordan normal form. The grey blocks

are called Jordan blocks.

In linear algebra, a Jordan normal form (often called

Jordan canonical form)[1] of a linear operator on a
nite-dimensional vector space is an upper triangular matrix of a particular form called a Jordan matrix, representing the operator with respect to some basis. Such matrix
has each non-zero o-diagonal entry equal to 1, immediately above the main diagonal (on the superdiagonal),
and with identical diagonal entries to the left and below
them. If the vector space is over a eld K, then a basis
with respect to which the matrix has the required form
exists if and only if all eigenvalues of the matrix lie in K,
or equivalently if the characteristic polynomial of the operator splits into linear factors over K. This condition is
always satised if K is the eld of complex numbers. The
diagonal entries of the normal form are the eigenvalues of
the operator, with the number of times each one occurs
being given by its algebraic multiplicity.[2][3][4]

1.2 Motivation
An n n matrix A is diagonalizable if and only if the
sum of the dimensions of the eigenspaces is n. Or,
equivalently, if and only if A has n linearly independent
eigenvectors. Not all matrices are diagonalizable. Consider the following matrix:

5
4
2
1
0
1 1 1

A=
.
1 1
3
0
1
1 1
2
Including multiplicity, the eigenvalues of A are = 1, 2,
4, 4. The dimension of the kernel of (A 4I ) is 1 (and
not 2), so A is not diagonalizable. However, there is an
invertible matrix P such that A = PJP 1 , where

If the operator is originally given by a square matrix M,

then its Jordan normal form is also called the Jordan normal form of M. Any square matrix has a Jordan normal
form if the eld of coecients is extended to one containing all the eigenvalues of the matrix. In spite of its name,
the normal form for a given M is not entirely unique, as it
is a block diagonal matrix formed of Jordan blocks, the
order of which is not xed; it is conventional to group
blocks for the same eigenvalue together, but no ordering
is imposed among the eigenvalues, nor among the blocks
for a given eigenvalue, although the latter could for instance be ordered by weakly decreasing size.[2][3][4]

1
0

J =
0
0

0
2
0
0

0
0
4
0

0
0

.
1
4

The matrix J is almost diagonal. This is the Jordan normal

form of A. The section Example below lls in the details
of the computation.
1

2 COMPLEX MATRICES

Complex matrices

2.1 Generalized eigenvectors

In general, a square complex matrix A is similar to a block Main article: Generalized eigenvectors
diagonal matrix
Consider the matrix A from the example in the previous
section. The Jordan normal form is obtained by some

J1
similarity transformation P 1 AP = J, i.e.

.
..
J =

Jp
AP = P J.
where each block J is a square matrix of the form

Ji =

1
i

..

..

1
i

Let P have column vectors pi, i = 1, ..., 4, then

[
A p1

p2

p3

] [
p4 = p1

p2

p3

1
]0
p4
0
0

0
2
0
0

0
0
4
0

So there exists an invertible matrix P such that P1 AP =

We see that
J is such that the only non-zero entries of J are on the
diagonal and the superdiagonal. J is called the Jordan
normal form of A. Each Ji is called a Jordan block of A.
In a given Jordan block, every entry on the superdiagonal (A 1I)p1 = 0
is 1.
(A 2I)p2 = 0
Assuming this result, we can deduce the following properties:
(A 4I)p3 = 0
Counting multiplicity, the eigenvalues of J, therefore (A 4I)p4 = p3 .
A, are the diagonal entries.
For i = 1,2,3 we have pi Ker(A i I) , i.e. p is an
Given an eigenvalue i, its geometric multiplicity is eigenvector of A corresponding to the eigenvalue . For
the dimension of Ker(A i I), and it is the number i=4, multiplying both sides by (A 4I) gives
of Jordan blocks corresponding to i.[10]
The sum of the sizes of all Jordan blocks corre(A 4I)2 p4 = (A 4I)p3 .
sponding to an eigenvalue i is its algebraic multiplicity.[10]
But (A 4I)p3 = 0 , so
A is diagonalizable if and only if, for every eigenvalue of A, its geometric and algebraic multiplicities coincide.
(A 4I)2 p4 = 0.
The Jordan block corresponding to is of the form
I + N, where N is a nilpotent matrix dened as
Nij = i,j (where is the Kronecker delta). The
nilpotency of N can be exploited when calculating
f(A) where f is a complex analytic function. For
example, in principle the Jordan form could give a
closed-form expression for the exponential exp(A).

Thus, p4 Ker(A 4I)2 .

Vectors such as p4 are called generalized eigenvectors of
A.

Thus, given an eigenvalue , its corresponding Jordan

block gives rise to a Jordan chain. The generator, or
lead vector, say pr, of the chain is a generalized eigenr
is the size of the
The number of Jordan blocks corresponding to of vector such that (A I) pr = 0, where rr1
Jordan
block.
The
vector
p
=
(A

I)
pr is an eigenj
j-1
1
size at least j is dim Ker(A - I) - dim Ker(A - I) .
vector
corresponding
to
.
In
general,
pi
is
a preimage of
Thus, the number of Jordan blocks of size exactly j
pi
under
A

I.
So
the
lead
vector
generates
the chain
is
via multiplication by (A I).
the statement that every square matrix A can
2 dim ker(Ai I)j dim ker(Ai I)j+1 dim ker(AiTherefore,
I)j1
be put in Jordan normal form is equivalent to the claim
Given an eigenvalue i, its multiplicity in the mini- that there exists a basis consisting only of eigenvectors
mal polynomial is the size of its largest Jordan block. and generalized eigenvectors of A.

0
[
0
= p1
1
4

2p2

2.3

Uniqueness

2.2

A proof

We give a proof by induction. The 1 1 case is trivial.

Let A be an n n matrix. Take any eigenvalue of A. The
range of A I, denoted by Ran(A I), is an invariant
subspace of A. Also, since is an eigenvalue of A, the
dimension Ran(A I), r, is strictly less than n. Let A'
denote the restriction of A to Ran(A I), By inductive
hypothesis, there exists a basis {p1 , ..., pr} such that A' ,
expressed with respect to this basis, is in Jordan normal
form.

2.3 Uniqueness
It can be shown that the Jordan normal form of a given
matrix A is unique up to the order of the Jordan blocks.

Next consider the subspace Ker(A I). If

Knowing the algebraic and geometric multiplicities of the

eigenvalues is not sucient to determine the Jordan normal form of A. Assuming the algebraic multiplicity m()
of an eigenvalue is known, the structure of the Jordan
form can be ascertained by analyzing the ranks of the
powers (A I)m() . To see this, suppose an n n matrix
A has only one eigenvalue . So m() = n. The smallest
integer k1 such that

Ran(A I) Ker(A I) = {0},

(A I)k1 = 0

the desired result follows immediately from the rank

is the size of the largest Jordan block in the Jordan form
nullity theorem. This would be the case, for example,
of A. (This number k1 is also called the index of . See
if A was Hermitian.
discussion in a following section.) The rank of
Otherwise, if
Q = Ran(A I) Ker(A I) = {0},
let the dimension of Q be s r. Each vector in Q is an
eigenvector of A' corresponding to eigenvalue . So the
Jordan form of A' must contain s Jordan chains corresponding to s linearly independent eigenvectors. So the
basis {p1 , ..., pr} must contain s vectors, say {prs, ...,
pr}, that are lead vectors in these Jordan chains from the
Jordan normal form of A'. We can extend the chains by
taking the preimages of these lead vectors. (This is the
key step of argument; in general, generalized eigenvectors need not lie in Ran(A I).) Let qi be such that
(A I)qi = pi for i = r s + 1, . . . , r.
Clearly no non-trivial linear combination of the qi can lie
in Ker(A I). Furthermore, no non-trivial linear combination of the qi can be in Ran(A I), for that would
contradict the assumption that each pi is a lead vector in
a Jordan chain. The set {qi}, being preimages of the linearly independent set {pi} under A I, is also linearly
independent.
Finally, we can pick any linearly independent set {z1, ...,
zt} that spans

(A I)k1 1
is the number of Jordan blocks of size k1 . Similarly, the
rank of

(A I)k1 2
is twice the number of Jordan blocks of size k1 plus the
number of Jordan bl Jordan structure of A. The general
case is similar.
This can be used to show the uniqueness of the Jordan
form. Let J 1 and J 2 be two Jordan normal forms of A.
Then J 1 and J 2 are similar and have the same spectrum,
including algebraic multiplicities of the eigenvalues. The
procedure outlined in the previous paragraph can be used
to determine the structure of these matrices. Since the
rank of a matrix is preserved by similarity transformation,
there is a bijection between the Jordan blocks of J 1 and
J 2 . This proves the uniqueness part of the statement.

3 Real matrices

If A is a real matrix, its Jordan form can still be non-real,

however there exists a real invertible matrix P such that
P1 AP = J is a real block diagonal matrix with each block
Ker(A I)/Q.
being a real Jordan block. A real Jordan block is either
By construction, the union the three sets {p1 , ..., pr}, identical to a complex Jordan block (if the corresponding
{qrs , ..., qr}, and {z1 , ..., zt} is linearly independent. eigenvalue i is real), or is a block matrix itself, consisting
Each vector in the union is either an eigenvector or a gen- of 22 blocks as follows (for non-real eigenvalue i =
eralized eigenvector of A. Finally, by ranknullity theo- ai + ibi ). The diagonal blocks are identical, of the form
rem, the cardinality of the union is n. In other words, we
have found a basis that consists of eigenvectors and gen[
]
eralized eigenvectors of A, and this shows A can be put
ai bi
Ci =
in Jordan normal form.
bi ai

CONSEQUENCES

and describe multiplication by i in the complex plane. ideal domain of polynomials with complex coecients.
The superdiagonal blocks are 22 identity matrices. The The monic element that generates I is precisely P.
full real Jordan block is given by
Let 1 , ..., q be the distinct eigenvalues of A, and si be
the size of the largest Jordan block corresponding to i.
It is clear from the Jordan normal form that the minimal

Ci I
polynomial of A has degree si.

..

.
Ci
While the Jordan normal form determines the minimal
.
Ji =

.
polynomial, the converse is not true. This leads to the
.

. I
notion of elementary divisors. The elementary divisors
Ci
of a square matrix A are the characteristic polynomials
This real Jordan form is a consequence of the complex of its Jordan blocks. The factors of the minimal polynoJordan form. For a real matrix the nonreal eigenvec- mial m are the elementary divisors of the largest degree
tors and generalized eigenvectors can always be chosen corresponding to distinct eigenvalues.
to form complex conjugate pairs. Taking the real and The degree of an elementary divisor is the size of the corimaginary part (linear combination of the vector and its responding Jordan block, therefore the dimension of the
conjugate), the matrix has this form with respect to the corresponding invariant subspace. If all elementary divinew basis.
sors are linear, A is diagonalizable.

Consequences

One can see that the Jordan normal form is essentially a

classication result for square matrices, and as such several important results from linear algebra can be viewed
as its consequences.

4.1

4.4 Invariant subspace decompositions

The Jordan form of a n n matrix A is block diagonal, and
therefore gives a decomposition of the n dimensional Euclidean space into invariant subspaces of A. Every Jordan
block Ji corresponds to an invariant subspace Xi. Symbolically, we put

Spectral mapping theorem

Cn =

Xi
Using the Jordan normal form, direct calculation gives a
i=1
spectral mapping theorem for the polynomial functional
calculus: Let A be an n n matrix with eigenvalues 1 , where each Xi is the span of the corresponding Jordan
..., n, then for any polynomial p, p(A) has eigenvalues chain, and k is the number of Jordan chains.
p(1 ), ..., p(n).
One can also obtain a slightly dierent decomposition via
the Jordan form. Given an eigenvalue i, the size of its
largest corresponding Jordan block s is called the index
4.2 CayleyHamilton theorem
of i and denoted by (i). (Therefore the degree of the
minimal polynomial is the sum of all indices.) Dene a
The CayleyHamilton theorem asserts that every ma- subspace Yi by
trix A satises its characteristic equation: if p is the
characteristic polynomial of A, then p(A) = 0. This can
be shown via direct calculation in the Jordan form, since Y = Ker( I A)(i ) .
i
i
any Jordan block for is annihilated by (X )m where m
is the multiplicity of the root of p, the sum of the sizes This gives the decomposition
of the Jordan blocks for , and therefore no less than the
size of the block in question. The Jordan form can be asl
sumed to exist over a eld extending the base eld of the

matrix, for instance over the splitting eld of p; this eld Cn =

Yi
i=1
extension does not change the matrix p(A) in any way.
where l is the number of distinct eigenvalues of A. Intuitively, we glob together the Jordan block invariant sub4.3 Minimal polynomial
spaces corresponding to the same eigenvalue. In the exThe minimal polynomial P of a square matrix A is the treme case where A is a multiple of the identity matrix
unique monic polynomial of least degree, m, such that we have k = n and l = 1.
P(A) = 0. Alternatively, the set of polynomials that an- The projection onto Yi and along all the other Yj ( j i )
nihilate a given A form an ideal I in C[x], the principal is called the spectral projection of A at i and is usually

5.2

Compact operators

denoted by P(i ; A). Spectral projections are mutually

orthogonal in the sense that P(i ; A) P(j ; A) = 0 if i j.
Also they commute with A and their sum is the identity
matrix. Replacing every i in the Jordan matrix J by one
and zeroising all other entries gives P(i ; J), moreover if
U J U 1 is the similarity transformation such that A = U
J U 1 then P(i ; A) = U P(i ; J) U 1 . They are not conned to nite dimensions. See below for their application
to compact operators, and in holomorphic functional calculus for a more general discussion.

5.2 Compact operators

In a dierent direction, for compact operators on a
Banach space, a result analogous to the Jordan normal
form holds. One restricts to compact operators because
every point x in the spectrum of a compact operator T,
the only exception being when x is the limit point of the
spectrum, is an eigenvalue. This is not true for bounded
operators in general. To give some idea of this generalization, we rst reformulate the Jordan decomposition in
the language of functional analysis.

Comparing the two decompositions, notice that, in general, l k. When A is normal, the subspaces Xi's in the
rst decomposition are one-dimensional and mutually or- 5.2.1 Holomorphic functional calculus
thogonal. This is the spectral theorem for normal operators. The second decomposition generalizes more easily For more details on this topic, see holomorphic funcfor general compact operators on Banach spaces.
tional calculus.
It might be of interest here to note some properties of the
index, (). More generally, for a complex number , its Let X be a Banach space, L(X) be the bounded operators
index can be dened as the least non-negative integer () on X, and (T) denote the spectrum of T L(X). The
such that
holomorphic functional calculus is dened as follows:

Fix a bounded operator T. Consider the family Hol(T) of

complex functions that is holomorphic on some open set
Ker( A)() = Ker( A)m , m ().
G containing (T). Let = {i} be a nite collection of
Jordan curves such that (T) lies in the inside of , we
So () > 0 if and only if is an eigenvalue of A. In the dene f(T) by
nite-dimensional case, () the algebraic multiplicity
of .

1
f (T ) =
f (z)(z T )1 dz.
2i

Generalizations

The open set G could vary with f and need not be connected. The integral is dened as the limit of the Riemann
sums, as in the scalar case. Although the integral makes
5.1 Matrices with entries in a eld
sense for continuous f, we restrict to holomorphic functions to apply the machinery from classical function theJordan reduction can be extended to any square maory (e.g. the Cauchy integral formula). The assumption
trix M whose entries lie in a eld K. The result states
that (T) lie in the inside of ensures f(T) is well dened;
that any M can be written as a sum D + N where D
it does not depend on the choice of . The functional calis semisimple, N is nilpotent, and DN = ND. This is
culus is the mapping from Hol(T) to L(X) given by
called the JordanChevalley decomposition. Whenever
K contains the eigenvalues of M, in particular when K is
algebraically closed, the normal form can be expressed
(f ) = f (T ).
explicitly as the direct sum of Jordan blocks.
Similar to the case when K is the complex numbers, We will require the following properties of this functional
knowing the dimensions of the kernels of (M I)k for calculus:
1 k m, where m is the algebraic multiplicity of the
eigenvalue , allows one to determine the Jordan form of
1. extends the polynomial functional calculus.
M. We may view the underlying vector space V as a K[x]module by regarding the action of x on V as application
2. The spectral mapping theorem holds: (f(T)) =
of M and extending by K-linearity. Then the polynomials
f((T)).
(x )k are the elementary divisors of M, and the Jordan
3. is an algebra homomorphism.
normal form is concerned with representing M in terms
of blocks associated to the elementary divisors.
The proof of the Jordan normal form is usually carried 5.2.2 The nite-dimensional case
out as an application to the ring K[x] of the structure theorem for nitely generated modules over a principal ideal In the nite-dimensional case, (T) = {i} is a nite disdomain, of which it is a corollary.
crete set in the complex plane. Let ei be the function that

EXAMPLE

is 1 in some open neighborhood of i and 0 elsewhere. 5.2.3 Poles of an operator

By property 3 of the functional calculus, the operator
Let T be a bounded operator be an isolated point of
(T). (As stated above, when T is compact, every point
in its spectrum is an isolated point, except possibly the
ei (T )
limit point 0.)
is a projection. Moreoever, let i be the index of i and
The point is called a pole of operator T with order if
the resolvent function RT dened by
f (z) = (z i )i .
The spectral mapping theorem tells us

RT () = ( T )1
has a pole of order at .

f (T )ei (T ) = (T i )i ei (T )

We will show that, in the nite-dimensional case, the orhas spectrum {0}. By property 1, f(T) can be directly der of an eigenvalue coincides with its index. The result
computed in the Jordan form, and by inspection, we see also holds for compact operators.
that the operator f(T)ei(T) is the zero matrix.
Consider the annular region A centered at the eigenvalue

with suciently small radius such that the intersecBy property 3, f(T) ei(T) = ei(T) f(T). So ei(T) is pretion
of the open disc B() and (T) is {}. The resolcisely the projection onto the subspace
vent function RT is holomorphic on A. Extending a result
from classical function theory, RT has a Laurent series
representation on A:
Ran e (T ) = Ker(T )i .
i

The relation

RT (z) =

am ( z)m

ei = 1

where

implies

Cn =

Ran ei (T ) =

Ker(T i )i

where the index i runs through the distinct eigenvalues of

T. This is exactly the invariant subspace decomposition

Cn =

Yi

1
am = 2i
( z)m1 (z T )1 dz and
C
C is a small circle centered at .
By the previous discussion on the functional calculus,
am = ( T )m1 e (T ) where e is 1
on B () and 0 elsewhere.
But we have shown that the smallest positive integer m
such that

given in a previous section. Each ei(T) is the projection

onto the subspace spanned by the Jordan chains corream = 0 and al = 0 l m
sponding to i and along the subspaces spanned by the
Jordan chains corresponding to j for j i. In other words
ei(T) = P(i;T). This explicit identication of the opera- is precisely the index of , (). In other words, the functors ei(T) in turn gives an explicit form of holomorphic tion RT has a pole of order () at .
functional calculus for matrices:
For all f Hol(T),

i 1

f (k)
(T i )k ei (T ).
k!

6 Example

This example shows how to calculate the Jordan normal

form of a given matrix. As the next section explains, it is
Notice that the expression of f(T) is a nite sum because, important to do the computation exactly instead of roundon each neighborhood of i, we have chosen the Taylor ing the results.
series expansion of f centered at i.
Consider the matrix
f (T ) =

i (T ) k=0

5
4
2
0
1
1
A=
1 1 3
1
1 1

0
P 1 AP = J =
0
0

1
1

0
2

which is mentioned in the beginning of the article.

The characteristic polynomial of A is

0
2
0
0

0
0
4
0

0
0
.
1
4

If we had interchanged the order of which the chain vectors appeared, that is, changing the order of v, w and {x,
y} together, the Jordan blocks would be interchanged.
However, the Jordan forms are equivalent Jordan forms.

() = det(IA) = 4 113 +422 64+32 = (1)(2)(4)2 .

This shows that the eigenvalues are 1, 2, 4 and 4, according to algebraic multiplicity. The eigenspace corresponding to the eigenvalue 1 can be found by solving the equation Av = v. It is spanned by the column vector v = (1,
1, 0, 0)T . Similarly, the eigenspace corresponding to the
eigenvalue 2 is spanned by w = (1, 1, 0, 1)T . Finally,
the eigenspace corresponding to the eigenvalue 4 is also
one-dimensional (even though this is a double eigenvalue)
and is spanned by x = (1, 0, 1, 1)T . So, the geometric
multiplicity (i.e., the dimension of the eigenspace of the
given eigenvalue) of each of the three eigenvalues is one.
Therefore, the two eigenvalues equal to 4 correspond to
a single Jordan block, and the Jordan normal form of the
matrix A is the direct sum

1
0
J = J1 (1) J1 (2) J2 (4) =
0
0

0
2
0
0

0
0
.
1
4

0
0
4
0

7 Numerical analysis
If the matrix A has multiple eigenvalues, or is close to
a matrix with multiple eigenvalues, then its Jordan normal form is very sensitive to perturbations. Consider for
instance the matrix
[
A=

]
1
.
1

If = 0, then the Jordan normal form is simply

[

]
1 1
.
0 1

However, for 0, the Jordan normal form is

[

1+
0
.
0
1

There are three chains. Two have length one: {v} and
{w}, corresponding to the eigenvalues 1 and 2, respec- This ill conditioning makes it very hard to develop a rotively. There is one chain of length two corresponding to bust numerical algorithm for the Jordan normal form, as
the result depends critically on whether two eigenvalues
the eigenvalue 4. To nd this chain, calculate
are deemed to be equal. For this reason, the Jordan normal form is usually avoided in numerical analysis; the sta
ble
Schur decomposition[11] or pseudospectra[12] are bet1
1

ter alternatives.
0 0
2
ker (A 4I) = span
0, 1 .

0
1

8 Powers

Pick a vector in the above span that is not in the kernel of

A 4I, e.g., y = (1,0,0,0)T . Now, (A 4I)y = x and (A
th
4I)x = 0, so {y, x} is a chain of length two corresponding If n is a natural number, the n power of a matrix in Jordan normal form will be a direct sum of upper triangular
to the eigenvalue 4.
matrices, as a result of block multiplication. More specifThe transition matrix P such that P 1 AP = J is formed by ically, after exponentiation each Jordan block will be an
putting these vectors next to each other as follows
upper triangular block.

1 1
1 1
[    ] 1 1 0 0
  
.
P = vwxy =
0
0 1 0
0
1
1 0

For example,

2
0

A computation shows that the equation P 1 AP = J indeed 0

holds.
0

1
2
0
0
0

0
1
2
0
0

0
0
0
5
0

4
0
16 32 24
0
0 16 32
0
0

0
0
= 0 0 16
0 0
1
0 625
5
0 0
0
0

0
0

0
.
500
625

12

Further, each triangular

block will consist of n on the
(n)
main diagonal, 1 times n1 on the upper diagonal, and
so on. This expression is valid for negative integer powers
as well if one (extends
)k (the)notion of the binomial coe(n)
|n|
n
cients k 7 |n|
k .

11 References
Beauregard, Raymond A.; Fraleigh, John B. (1973),
A First Course In Linear Algebra: with Optional
Introduction to Groups, Rings, and Fields, Boston:
Houghton Miin Co., ISBN 0-395-14017-X

For example,

1
0

0
0

1
1
0
0
0

0
1
1
0
0

0
0
0
2
0

n n
1
0
0
0

0
=0
0

1
2
0

(n) n1
1 1
n1
0
0
0

(n) n2
1
(n2 )n1
1 1
n1
0
0

See also
Canonical form
Frobenius normal form
Jordan matrix
JordanChevalley decomposition
Matrix decomposition
Weyr canonical form

10

Notes

[1] Shilov denes the term Jordan canonical form and in a

footnote says that Jordan normal form is synonymous.
These terms are sometimes shortened to Jordan form.
(Shilov) The term Classical canonical form is also sometimes used in the sense of this article. (James & James,
1976)
[2] Beauregard & Fraleigh (1973, pp. 310316)
[3] Golub & Van Loan (1996, p. 355)
[4] Nering (1970, pp. 118127)
[5] Beauregard & Fraleigh (1973, pp. 270274)
[6] Golub & Van Loan (1996, p. 353)
[7] Nering (1970, pp. 113118)
[8] Cullen (1966, p. 114)

EXTERNAL LINKS

Cullen, Charles G. (1966), Matrices and Linear Transformations, Reading: Addison-Wesley,

LCCN
0
066021267

0
0
N. Dunford
J.T. Schwartz, Linear Operators,
and
.
0 ( )0

Part
I:
General
Theory,
Interscience, 1958.
n n1
n2
1 2
n
0 DanielT.
2 Finkbeiner II, Introduction to Matrices and
Linear Transformations, Third Edition, Freeman,
1978.
Franklin, Joel N. (1968), Matrix Theory, Englewood
Clis: Prentice-Hall, LCCN 68016345
Gene H. Golub and Charles F. Van Loan, Matrix
Computations (4th ed.), Johns Hopkins University
Press, Baltimore, 2012.
Gene H. Golub and J. H. Wilkinson, Ill-conditioned
eigensystems and the computation of the Jordan normal form, SIAM Review, vol. 18, nr. 4, pp. 578
619, 1976.
Horn, Roger A.; Johnson, Charles R. (1985), Matrix
Analysis, Cambridge University Press, ISBN 978-0521-38632-6.
Glenn James and Robert C. James, Mathematics
Dictionary, Fourth Edition, Van Nostrand Reinhold,
1976.
Saunders MacLane and Garrett Birkho, Algebra,
MacMillan, 1967.
Anthony N. Michel and Charles J. Herget, Applied
Algebra and Functional Analysis, Dover, 1993.
Nering, Evar D. (1970), Linear Algebra and Matrix Theory (2nd ed.), New York: Wiley, LCCN
76091646
Georgi E. Shilov, Linear Algebra, Dover, 1977.
I. R. Shafarevich & A. O. Remizov (2012) Linear
Algebra and Geometry, Springer ISBN 978-3-64230993-9.
Jordan Canonical
world.wolfram.com

Form

article

at

math-

[9] Franklin (1968, p. 122)

[10] Horn & Johnson (1985, 3.2.1)
[11] See Golub & Van Loan (2014), 7.6.5; or Golub &
Wilkinson (1976) for details.
[12] See Golub & Van Loan (2014), 7.9

12 External links
On line tool on Jordan form and matrix diagonalizations by www.mathstools.com

13
13.1

Text and image sources, contributors, and licenses

Text

Jordan normal form Source: http://en.wikipedia.org/wiki/Jordan%20normal%20form?oldid=656810162 Contributors: Michael Hardy,

Kku, Stan Shebs, Ideyal, Charles Matthews, Dysprosia, Jitse Niesen, David Shay, McKay, Robbot, MathMartin, Tobias Bergemann,
Tosha, Giftlite, BenFrantzDale, Waltpohl, Helohe, Mecanismo, Gauge, Rgdboer, Krakhan, David Crawshaw, HasharBot, Squizzz, Giant toaster, Derbeth, Florian Beissner, Natalya, Joriki, MattGiuca, MFH, Reedbeta, Heptor, Fresheneesz, Algebraist, YurikBot, Wavelength, Gaius Cornelius, Buster79, NickBush24, Taco325i, Crasshopper, Gadget850, Ms2ger, KnightRider, SmackBot, Spireguy, Chris the
speller, Nbarth, Foxjwill, MarkC77, WalterMB, Vaughan Pratt, CmdrObot, Mct mht, Thijs!bot, Salgueiro, GromXXVII, Liberio, Stdazi,
Xblin, Outlook, Jtir, Leyo, Aerolsh, Malkoun, Krishnachandranvn, Warut, LokiClock, Semidimes, TXiKiBoT, MRFS, Forwardmeasure,
JerroldPease-Atlanta, Denisarona, Mr. Granger, ClueBot, MATThematical, Mild Bill Hiccup, SetaLyas, 800km3rk, Bender2k14, Juancentro, Varj, Franklin.vp, Qwfp, Marc van Leeuwen, Fj217, Addbot, Luckas-bot, Yobot, Calle, APh8ohph, Ivan Kuckir, Xqbot, Maxwell
helper, FrescoBot, Kiefer.Wolfowitz, Georges.dib, MedicineMan555, Amirhdrz 91, Aadityakalsi, SamantaGhezzi, ZroBot, Quondum,
Anita5192, Wcherowi, Dr-er, Dexbot, AHusain314, Carlostp12, UtherSB and Anonymous: 99

13.2

Images

File:Jordan_blocks.svg Source: http://upload.wikimedia.org/wikipedia/commons/4/4f/Jordan_blocks.svg License: CC BY-SA 3.0 Contributors: Own work Original artist: Jakob.scholbach

13.3

Content license

Creative Commons Attribution-Share Alike 3.0