Matrices
Matrices
െ ሺܣሻߣ ȁܣȁ
ൌ ሺെͳሻ͵ ߣ͵ Ͳߣ െ ሺͷ െ െ ʹሻߣ ʹሾሺെͷሻ െ ʹሺͳሻሿ
ൌ െߣ͵ ͳ͵ߣ െ ͳʹ
(c) Characteristic equation is െߣ͵ ͳ͵ߣ െ ͳʹ ൌ Ͳ
ฺ ߣ͵ െ ͳ͵ߣ ͳʹ ൌ Ͳ
ฺ ߣ͵ െ ߣ െ ͳʹߣ ͳʹ ൌ Ͳ
ฺ ߣሺߣʹ െ ͳሻ െ ͳʹሺߣ െ ͳሻ ൌ Ͳ
ฺ ሺߣ െ ͳሻሾߣሺߣ ͳሻ െ ͳʹሿ ൌ Ͳ
ฺ ሺߣ െ ͳሻሾߣʹ ߣ െ ͳʹሿ ൌ Ͳ
ฺ ሺߣ െ ͳሻሺߣ Ͷሻሺߣ െ ͵ሻ ൌ Ͳ
ͳߣ ൌ ͳǡ ߣʹ ൌ െͶǡ and ߣ͵ ൌ ͵ are the characteristic roots of the matrix.
The matrix is non symmetric. Eigen values distinct.
2. Find the eigen values and corresponding eigen vectors of the matrix
ૡ െ
ൌ െ ૠ െ൩Ǥ
െ
Solution: The characteristic equation of the matrix ܣis given by ȁ ܣെ ߣܫȁ ൌ Ͳ
ͺെߣ െ ʹ
֜ อ െ െ ߣ െͶ อ ൌ Ͳ
ʹ െͶ ͵ െ ߣ
֜ െߣ ͳͺߣʹ െ Ͷͷߣ ൌ Ͳ
͵
֜ ߣ͵ െ ͳͺߣʹ Ͷͷߣ ൌ Ͳ
֜ ߣሺߣ െ ͵ሻሺߣ െ ͳͷሻ ൌ Ͳǡ ߣ ൌ Ͳǡ ͵ǡͳͷ
which are the required eigen values of ܣǤ
MATRIX THEORY | 3
ͳݔ
ݔ
We shall now find the eigen vector ܺ ൌ ʹ ൩ corresponding to each value of ߣ by considering
͵ݔ
the equation ሺ ܣെ ߣܫሻܺ ൌ Ͳ ǥ ǥ ǥ ǥ ǥ ǥ ǥ ǥ ǥ ǥ ǥ ǥ ǥ ሺͳሻ
For ߣ ൌ Ͳ, (1) gives
ͺ െ ʹ ͳݔ Ͳ
െ െͶ൩ ʹݔ൩ ൌ Ͳ൩
ʹ െͶ ͵ ͵ݔ Ͳ
ͺ െ ʹ ͳݔ Ͳ
̱ Ͳ െͷ ͷ൩ ʹݔ൩ ൌ Ͳ൩ ሺܴʹ ͵ܴ͵ ሻ
ʹ െͶ ͵ ͵ݔ Ͳ
֜ ͺ ͳݔെ ʹݔ ʹ ͵ݔൌ ͲǢെͷ ʹݔ ͷ ͵ݔൌ ͲǢ ʹ ͳݔെ Ͷ ʹݔ ͵ ͵ݔൌ Ͳ
ʹݔ ൌ ͵ݔ
Let ͵ݔൌ ݐ, where ݐis a parameter. Then ʹݔൌ ݐ
ͳ ͳ ݐ
ሾ ʹݔെ ʹ ͵ݔሿ ൌ ሾ ݐെ ʹݐሿ ൌ
ͳݔ ൌ
ͺ ͺ ʹ
ݐȀʹ ͳ
Hence ܺ ൌ ݐ൩ which may be represented by ܺͳ ൌ ʹ൩
ݐ ʹ
For ߣ ൌ ͵ǡ ሺͳሻ gives
ͷ െ ʹ ͳݔ Ͳ
െ Ͷ െͶ൩ ʹݔ൩ ൌ Ͳ൩
ʹ െͶ Ͳ ͵ݔ Ͳ
֜ ͷ ͳݔെ ʹݔ ʹ ͵ݔൌ ͲǢെ ͳݔ Ͷ ʹݔെ Ͷ ͵ݔൌ ͲǢ ʹ ͳݔെ Ͷ ʹݔൌ Ͳ
ͳݔ ൌ ʹʹݔ
Let ʹݔൌ ݐǡ then ͳݔൌ ʹݐ
ͳ ͳ
͵ݔ ൌ ሾെͷ ͳݔ ʹݔሿ ൌ ሾെͷሺʹݐሻ ݐሿ ൌ െʹݐ
ʹ ʹ
ʹݐ ʹ
Hence ܺ ൌ ݐ൩ which may be represented by ܺʹ ൌ ͳ ൩
െʹݐ െʹ
For ߣ ൌ ͳͷǡ ሺͳሻ gives
െ െ ʹ ͳݔ Ͳ
െ െͺ െͶ ൩ ʹݔ൩ ൌ Ͳ൩
ʹ െͶ െͳʹ ͵ݔ Ͳ
െ െ ʹ ݔ ͳ Ͳ
̱ Ͳ െʹͲ െͶͲ൩ ʹݔ൩ ൌ Ͳ൩ሺܴʹ ͵ܴʹ ሻ
ʹ െͶ െͳʹ ͵ݔ Ͳ
֜ െ ͳݔെ ʹݔ ʹ ͵ݔൌ ͲǢെʹͲ ʹݔെ ͶͲ ͵ݔൌ ͲǢ ʹ ͳݔെ Ͷ ʹݔെ ͳʹ ͵ݔൌ Ͳ
ʹݔ ൌ െʹ͵ݔ
4 | APPLIED MATHEMATI CS- III
ͳ െͳ ͳ ܺͳ Ͳ
(i) For ߣ ൌ ͳǡ ሾ ܣെ ߣͳ ܫሿܺ ൌ Ͳ gives ͳ ͳ െͳ൩ ܺʹ ൩ ൌ Ͳ൩
ͳ െͳ ͳ ܺ͵ Ͳ
We may use matrix method to obtain the roots of the above equation
ܴʹ െ ܴͳ ͳ െͳ ͳ ͳ Ͳ
By
ܴ͵ െ ܴͳ
Ͳ ʹ െʹ൩ ʹ ൩ ൌ Ͳ൩
Ͳ Ͳ Ͳ ͵ Ͳ
ͳ െ ʹ ͵ ൌ Ͳ
ʹʹ െ ʹ͵ ൌ Ͳ
i.e. ʹ െ ͵ ൌ Ͳ
We note that the rank of the matrix is 2 and the number of variables is 3
Hence, there is ͵ െ ʹ ൌ ͳliniarly independent solution.
Putting ܺ͵ ൌ ͳǡ ʹܺݐ݁݃݁ݓൌ ͳܽ݊݀ܺͳ ൌ Ͳ
Ͳ
Hence, corresponding to the Eigen value 1, the Eigen vector is ͳ൩or ሾͲǡ ͳǡ ͳሿԢ
ͳ
If K is a non-zero scale then k X also an Eigen vector.
Ͳ െͳ ͳ ͳ Ͳ
(ii) For ɉ ൌ ʹǡ ሾ െ ɉʹ ሿ ൌ Ͳ gives ͳ Ͳ െͳ൩ ʹ ൩ ൌ Ͳ൩
ͳ െͳ Ͳ ͵ Ͳ
െܺʹ ܺ͵ ൌ Ͳ െ ܺͳ െ ܺ͵ ൌ Ͳǡ ܺͳ െ ܺʹ ൌ Ͳ
If ܺ͵ ൌ ͳǡ ܺʹ ൌ ͳǡ ܺͳ ൌ ͳ
ͳ
Hence, corresponding to the Eigen value ߣ ൌ ʹ the Eigen vector is ͳ൩ or ሾͳǡ ͳǡ ͳሿԢ
ͳ
If k is non-zero scalar then kX also is an Eigen vector.
െͳ െͳ ͳ ܺͳ Ͳ
ሺܑܑܑሻ For ߣ ൌ ͵, ܣെ ߣ͵ ܫሿ ൌ Ͳ gives ͳ െͳ ܺ
െͳ൩ ʹ ൩ ൌ Ͳ൩
ͳ െͳ െͳ ܺ͵ Ͳ
ܴʹ ܴͳ െͳ െͳ ͳ ܺͳ Ͳ
By
ܴʹ
ܴ͵ Ͳ െʹ Ͳ൩ ܺʹ ൩ ൌ Ͳ൩
Ͳ Ͳ Ͳ ܺ͵ Ͳ
െͳ െ ʹ ͵ ൌ Ͳ
ʹ ʹ ൌ Ͳ
ʹ ൌ Ͳ
If ܺ͵ ൌ ͳܺͳ ൌ ͳ
MATRIX THEORY | 7
ͳ
Hence, corresponding to the Eigen value ߣ ൌ ͵. The Eigen vector is Ͳ൩or ሾͳǡ Ͳǡ ͳሿԢ
ͳ
If k is a non-zero scalar then kX also is an Eigen vector.
To verify the liner independence of vectors, consider
ͳܺ ͳܭ ݇ʹ ܺʹ ͵ܺ ͵ܭൌ Ͳ
ͳܭ ሾͲǡͳǡͳሿ ʹܭሾͳǡͳǡͳሿ ͵ܭሾͳǡͲǡͳሿ ൌ Ͳ
Ͳ ʹ ͵ ൌ Ͳ……………………(i)
ͳ ʹ Ͳ ൌ Ͳ……………………...(ii)
ͳ ʹ Ͳ ൌ Ͳ……………………..(iii)
From (i) and (iii), we get ͳܭൌ ͲǡFrom (ii) and (iii), we get ͵ܭൌ Ͳ
Hence, from (iii), we get ʹܭൌ Ͳ.
Hence ͳ ൌ Ͳǡ ʹ ൌ Ͳǡ ͵ ൌ Ͳ
The vectors are linearly independent (refer to the theorem)
െ
5. Find the Eigen values and Eigen vectors for the matrix െ െ൩
െ
͵ െ ɉ െͳ ͳ
Solution: The characteristic equation is െͳ ͷ െ ɉ െͳ ൩
ͳ െͳ ͵ െ ɉ
ሺ͵ െ ɉሻሾሺͷ െ ɉሻሺ͵ െ ɉሻ െ ͳሿ ͳሾെͳሺ͵ െ ɉሻ ͳሿ ͳሾͳ െ ሺͷ െ ɉሻ ൌ Ͳሿ
ɉ͵ െ ͳͳɉʹ ͵ɉ െ ͵ ൌ Ͳ
ߣ͵ െ ʹߣʹ െ ͻߣʹ െ ͳͺߣ ͳͺߣ െ ͵ ൌ Ͳ
ሺߣ െ ʹሻሺߣʹ െ ͻߣ ͳͺሻ ൌ Ͳ
ሺߣ െ ʹሻሺߣ െ ͵ሻሺߣ െ ሻ ൌ Ͳ
ߣ ൌ െʹǡ͵ǡ
ͳ െͳ ͳ ͳ Ͳ
i) ɉ െ ʹǡ ሾ െ ɉͳ ሿ ൌ Ͳ gives െͳ ͵ െͳ൩ ʹ ൩ ൌ Ͳ൩
ͳ െͳ ͳ ͵ Ͳ
ʹ ͳ ͳ െͳ ͳ ͳ Ͳ
By
͵ െ
ͳ Ͳ ʹ Ͳ൩ ʹ ൩ ൌ Ͳ൩
Ͳ Ͳ Ͳ ͵ Ͳ
ͳܺ െ ܺʹ ܺ͵ ൌ Ͳǡ ܺʹ ൌ Ͳ
ʹ ൌ Ͳǡ ͵ ൌ െͳͳ ൌ ͳ
8 | APPLIED MATHEMATI CS- III
Hence, corresponding to the Eigen value ߣ ൌ ʹǡ we get the following Eigen vector
ܺ ൌ ሾͳǡͲǡ െͳሿ
If K is non-zero scalar kX also is an Eigen vector
Ͳ െͳ ͳ ܺͳ Ͳ
ሻ ߣ ൌ ͵ǡ ሾ െ ɉʹ ሿ ൌ Ͳ gives െͳ ʹ െͳ൩ ܺʹ ൩ ൌ Ͳ൩
ͳ െͳ Ͳ ܺ͵ Ͳ
Ͳ െͳ ͳ ܺͳ Ͳ
By ܴʹ ܴͳ െͳ ͳ Ͳ൩ ܺʹ ൩ ൌ Ͳ൩
ͳ െͳ Ͳ ܺ͵ Ͳ
െܺʹ ܺ͵ ൌ Ͳǡ െܺͳ ܺʹ ൌ Ͳǡ ܺͳ െ ܺʹ ൌ Ͳ
Putting ܺ͵ ൌ ͳǡ ܺʹ ൌ ͳܺͳ ൌ ͳ
Hence, corresponding to the Eigen value ߣ ൌ ͵ǡ we get the following Eigen vector,
ܺ ൌ ሾͳǡͳǡͳሿԢ
If K is a non-zero scalar kX is also an Eigen vector.
െ͵ െͳ ͳ ܺͳ Ͳ
(iii) ߣ ൌ ǡ ሾ ܣെ ߣ͵ ሿ ൌ Ͳ gives െͳ െͳ െͳ൩ ܺʹ ൩ ൌ Ͳ൩
ͳ െͳ െ͵ ܺ͵ Ͳ
ܴʹ ܴͳ െ͵ െͳ ͳ ܺͳ Ͳ
By
ܴ͵
͵ܴͳ െͶ െʹ Ͳ൩ ܺʹ ൩ ൌ Ͳ൩
െͺ െͶ Ͳ ܺ͵ Ͳ
െ͵ܺͳ െ ܺʹ ܺ͵ ൌ Ͳܽ݊݀ െ ʹܺͳ െ ܺʹ ൌ Ͳ
Putting ܺͳ ൌ ͳ, we get ܺʹ ൌ െʹǡ ܺ͵ ൌ ͳǤ
Hence, corresponding to the Eigen value ߣ ൌ ǡwe get the following Eigen vector,
ܺ ൌ ሾͳǡ െʹǡͳሿԢ
If K is a non-zero scalar kX is an Eigen vector
Note: Any two Eigen vectors corresponding to two distinct Eigen values of a real symmetric
matrix are orthogonal.
6. Find the Eigen values and Eigen vectors for the matrix ൩
ͳെߣ ʹ ʹ
Solution: The characteristic equation is ʹ ͳെߣ ʹ ൩ൌͲ
ʹ ʹ ͳെߣ
ʹ
ሺͳ െ ߣሻሾሺͳ െ ߣሻ െ Ͷሿ െ ʹሾʹሺͳ െ ߣሻ െ Ͷሿ ʹሾͶ െ ʹሺͳ െ ߣሻሿ ൌ Ͳ
͵ߣ െ ͵ߣʹ െ ͻߣ െ ͷ ൌ Ͳ
͵ߣ ߣʹ െ Ͷߣʹ െ Ͷߣ െ ͷߣ ൌ Ͳ
MATRIX THEORY | 9
ሺߣ ͳሻሺߣʹ െ Ͷߣ െ ͷሻ ൌ Ͳ
ሺߣ ሺߣ ͳሻሺߣ ͳሻሺߣ െ ͷሻ ൌ Ͳ
ߣ ൌ െͳǡ െͳǡͷ
ʹ ʹ ʹ ܺͳ Ͳ
(ii) For ߣ ൌ െͳሾ ܣെ ߣʹ ܫሿܺ ൌ Ͳ gives ʹ ʹ ʹ൩ ܺʹ ൩ ൌ Ͳ൩
ʹ ʹ ʹ ܺ͵ Ͳ
ܴʹ െ ܴͳ ͳ ͳ ͳ ܺͳ Ͳ
By ܴ͵ െ ܴʹ Ͳ Ͳ Ͳ൩ ܺʹ ൩ ൌ Ͳ൩
ሺͳΤʹሻ ܴͳ Ͳ Ͳ Ͳ ܺ͵ Ͳ
The rank of the matrix is 1 and the number of variables is 3. Hence, there are ͵ െ ͳ ൌ ʹ linearly
independent solutions.
Putting ܺ͵ ൌ Ͳܽ݊݀ܺʹ ൌ െͳǡ ͳܺݐ݁݃݁ݓൌ ͳǤ
And putting ܺʹ ൌ Ͳܽ݊݀ܺʹ ൌ െͳǡ ͳܺݐ݁݃݁ݓൌ ͳ
Hence, ܺͳ ൌ ሾͳǡ െͳǡͲሿԢ ܽ݊݀ܺʹ ൌ ሾͳǡͲǡ െͳሿԢ
െͶ ʹ ʹ ܺͳ Ͳ
(iii) ɉ ൌ ͷǡ ሾ െ ߣʹ ܫሿܺ ൌ Ͳ gives ʹ െͶ ʹ ൩ ܺʹ ൩ ൌ Ͳ൩
ʹ ʹ െʹ ܺ͵ Ͳ
The first two rows give
Ͷܺͳ ʹܺʹ ʹܺ͵ ൌ Ͳ
ʹܺͳ െ Ͷܺʹ ʹܺ͵ ൌ Ͳ
ฺ െʹܺͳ ܺʹ ܺ͵ ൌ Ͳ
ܺͳ െ ʹܺʹ ܺ͵ ൌ Ͳ
By crammer’s rule
ͳݔ െʹݔ ͵ݔ
ൌ ൌ
ͳ ʹ െʹ െ ͳ Ͷ െ ͳ
͵ݔ ʹݔ ͳݔ
ൌ ൌ
͵ ͵ ͵
͵ݔ ʹݔ ͳݔ
ൌ ൌ
ͳ ͳ ͳ
Hence, corresponding to ߣ ൌ ͷ, the Eigen vector is ͵ݔൌ ሾͳǡͳǡͳሿ
Let us check whether the Eigen vectors corresponding to the equal Eigen value -1 are
independent. We have ܺͳ ൌ ሾͳǡ െͳǡͲሿԢ ܺʹ ൌ ሾͳǡͲǡͳሿ
Now, ݇ͳ ܺͳ ݇ʹ ܺʹ ൌ Ͳ
ͳ݇ ሾͳǡ െͳǡͲሿ ݇ʹ ሾͳǡͲǡ െͳሿ ൌ Ͳ
ͳ݇ ݇ʹ ൌ Ͳ െ ݇ͳ ൌ Ͳǡ െ݇ʹ ൌ Ͳ
Hence the vectors are linearly independent
10 | APPLIED MATHEMATICS - III
െ
7. Find the eigen values and eigen vectors of the matrix ൩.
െ
ʹ െʹ ͵ ͳ Ͳ Ͳ
Solution: The characteristic matrix is ܣെ ߣ ܫൌ ͳ ͳ ͳ ൩ െ ߣ Ͳ ͳ Ͳ൩
ͳ ͵ െͳ Ͳ Ͳ ͳ
ʹെߣ െʹ ͵
ൌ ͳ ͳെߣ ͳ ൩
ͳ ͵ െͳ െ ߣ
ʹെߣ െʹ ͵
The characteristic polynomial is ȁ ܣെ ߣܫȁ ൌ ͳ ͳെߣ ͳ ൩
ͳ ͵ െͳ െ ߣ
͵ ͵ ʹ
ൌ ሺെͳሻ ߣ ʹߣ ͷߣ െ
Characteristic equation is
െߣ͵ ʹߣʹ ͷߣ െ ൌ Ͳ
ฺ ߣ͵ െ ʹߣʹ െ ͷߣ ൌ Ͳ
On factorization we get ߣ ൌ െʹ, ߣ ൌ ͳ, ߣ ൌ ͵.
The characteristic vector has to satisfy the equation ሺ ܣെ ߣܫሻܺ ൌ Ͳ which for given matrix ܣis
ʹെߣ െʹ ͵ ͳݔ Ͳ
ͳ ͳെߣ ݔ
ͳ ൩ ʹ ൩ ൌ Ͳ൩
ͳ ͵ െͳ െ ߣ ͵ݔ Ͳ
Corresponding to ߣ ൌ െʹ the eigen vector ߣͳ is given by
Ͷ െʹ ͵ ͳݔ Ͳ
ͳ ͵ ͳ൩ ʹݔ൩ ൌ Ͳ൩
ͳ ͵ ͳ ͵ݔ Ͳ
To obtain rank of ܣെ ߣͳ ܫwe have
Ͷ
െʹ ͵
ܣെ ߣͳ ܫൌ ͳ
͵ ͳ൩
ͳ
͵ ͳ
Ͷ െʹ ͵
ܴ͵ െ ܴʹ ̱݃݅ݏ݁ݒͳ ͵ ͳ൩
Ͳ Ͳ Ͳ
Solving the equations by Cramer’s rule
Ͷ ͳݔെ ʹ ʹݔ ͵ ͵ݔൌ Ͳ
ͳݔ ͵ ʹݔ ͵ݔൌ Ͳ
ͳݔ െʹݔ ͵ݔ
ൌ ൌ
െʹ ͵ Ͷ ͵ Ͷ െʹ
ቚ ቚ ቚ ቚ ቚ ቚ
͵ ͳ ͳ ͳ ͳ ͵
MATRIX THEORY | 11
Theorem: 3. Show that characteristic roots of a diagonal matrix are principal diagonal elements
of a matrix.
Proof: Consider a diagonal matrix ܣis given by
ܽͳͳ Ͳ Ͳ
ܣൌ Ͳ ܽʹʹ Ͳ ൩
Ͳ Ͳ ܽ͵͵
The characteristic equation of ܣis given by ȁ ܣെ ߣܫȁ ൌ Ͳ
ܽͳͳ െ ߣ Ͳ Ͳ
ȁ ܣെ ߣܫȁ ൌ อ Ͳ ܽʹʹ െ ߣ Ͳ อൌͲ
Ͳ Ͳ ܽ͵͵ െ ߣ
MATRIX THEORY | 15
Theorem: 5. If ߣ is eigen value of matrix ܣthen show that ߣ݊ is the eigen value of ݊ܣ.
Proof: Let ԢܣԢ be the square matrix. ߣ be the eigen value of ܣand ܺ the corresponding eigen
vector of ܣ.
By definition
ܺܣൌ ߣܺ ǥ ǥ ǥ ǥ ሺͳሻ
Multiplying equation (1) by ܣ.
ܣሺܺܣሻ ൌ ܣሺߣܺሻ
ฺ ܺ ʹܣൌ ߣሺܺܣሻ
ฺ ܺ ʹܣൌ ߣሺߣܺሻሾ ܺܣൌ ߣܺሺͳሻሿ
ฺ ܺ ʹܣൌ ߣʹ ܺ ǥ ǥ ǥ ǥ ǥ ǥ ሺʹሻ
Here ߣʹ is the eigen value of ʹܣ.
Multiplying equation (2) by ܣ.
ܣሺܺ ʹܣሻ ൌ ܣሺߣʹ ܺሻ
16 | APPLIED MATHEMATICS - III
ฺ ܺ ͵ܣൌ ߣʹ ሺܺܣሻ
ฺ ܺ ͵ܣൌ ߣʹ ሺߣܺሻ
ฺ ܺ ͵ܣൌ ߣ͵ ܺ
Here ߣ͵ is the eigen value of ͵ܣ.
In general ߣ݊ is the eigen value of ݊ܣ
Theorem: 6. Show that the value of determinant of an orthogonal matrix is േͳ.
Proof: let ܣbe an orthogonal matrix.
By Definition ܣܣԢ ൌ ܣԢ ܣൌ ͳ
ȁܣܣԢȁ ൌ ȁܫȁ
ฺ ȁܣȁȁܣԢȁ ൌ ͳ
ฺ ȁܣȁȁܣȁ ൌ ͳሾȁܣԢȁ ൌ ȁܣȁሿ
ฺ ȁܣȁʹ ൌ ͳ
ฺ ȁܣȁ ൌ േͳ
Value of determinant of an orthogonal matrix is േͳ.
Theorem: 7. Show that the characteristic roots of an orthogonal matrix are of unit modulus.
Proof: Let ܣbe the given matrix which is orthogonal.
ฺ ܣԢ ܣൌ ܣܣԢ ൌ ܫ
Let ߣ be the characteristic root of ܣand ܺ be the characteristic vector of ܣ.
By definition
ܺܣൌ ߣܺ ǥ ǥ ǥ ǥ ሺͳሻ
ฺ ሺܺܣሻԢ ൌ ሺߣܺሻԢ
ฺ ܺ Ԣ ܣԢ ൌ ߣܺԢ
On multiplying both sides by ܺܣൌ ߣܺ
ܺ ԢܣԢሺܺܣሻ ൌ ߣܺ Ԣ ሺߣܺሻ
ฺ ܺ Ԣ ሺܣԢܣሻܺ ൌ ߣʹ ሺܺԢܺሻ
ฺ ܺ Ԣ ሺܫሻܺ ൌ ߣʹ ሺܺԢܺሻ
ฺ ܺԢܺ ൌ ߣʹ ሺܺԢܺሻ
ฺ ͳ ൌ ߣʹ
ฺ ߣ ൌ േͳ
ȁߣȁ ൌ ͳ
Characteristic roots of an orthogonal matrix are of unit modulus.
MATRIX THEORY | 17
SOLVED PROBLEMS
ሺሻ െሺെሻ
1. Prove that the Eigen values of ሺሻ ሺെሻ
are of unit modulus.
൫ͳ േ ݅ξ͵൯ ൫ͳ െ ݅ξ͵൯ ͳ െ ͵݅ ʹ
ߣߣҧ ൌ ൌ ൌͲ
ʹ ʹ Ͷ
Hence, Eigen values are of unit modulus.
2. Prove that similar matrices have same character roots hence or otherwise find
whether & are similar matrices.
Solution: LetܣƬܤare similar matrices.
a non-singular matrix, ܯsuch that
ܤൌ ܯെͳ ܯܣ
ܤെ ɉ ൌ ܯെͳ ܯܣെ ɉ
ܤെ ɉ ൌ ܯെͳ ܯܣെ ܯെͳ ɉ
ൌ ܯെͳ ሺ ܣെ ɉ ሻܯ
݀݁ݐሺ ܤെ ɉ ሻ ൌ ܯെͳ Ǥ ݀݁ݐሺ ܣെ ɉ ሻǤ ݀݁ܯݐ
ൌ ܯെͳ Ǥ ݀݁ܯݐǤ ݀݁ݐሺ ܣെ ɉ ሻ
ൌ ሺܯെͳ ܯሻ Ǥ ሺ ܣെ ɉ ሻ༌
ൌ ͳǤ ݀݁ݐሺ ܣെ ɉ ሻ
18 | APPLIED MATHEMATICS - III
Thus ܤ&ܣhave same characteristic polynomials and hence same characteristic roots
ʹ ͳ ͵ Ͳ
Given ܣൌ ቂ ቃ ܽ݊݀ ܤൌ ቂ ቃ
Ͳ ͳ ͳ ʹ
ሺ݅ሻሾ ܣെ ɉ ሿܺ ൌ Ͳ
ʹ െ ɉ ͳ ͳݔ Ͳ
ቂ ቃቂ ݔቃ ൌ ቂ ቃ
Ͳ ͳ െ ɉ ʹ Ͳ
ሺ݅݅ሻȁ ܣെ ɉ ȁ ൌ Ͳ
ʹ െ ɉ ͳ
ቚ ቚൌ Ͳ
Ͳ ͳ െ ɉ
ɉʹ െ ͵ɉ ʹ ൌ Ͳ
ฺ ɉ ൌ ʹǡ ͳ
Alsoǡ ሾ ܤെ ɉ ሿܺ ൌ Ͳǡ
͵െɉ Ͳ ͳݔ Ͳ
ቂ ቃቂ ݔቃ=ቂ ቃ
ͳ ʹെɉ ʹ Ͳ
ȁ ܤെ ɉ ȁ ൌ Ͳ
͵െɉ Ͳ
ቚ ቚൌͲ
ͳ ʹെɉ
ɉʹ െ ͷɉ ൌ Ͳ
ฺ ɉ ൌ ͵ǡ ʹ
Since, characteristic roots ofܤ&ܣare not same.
ܣ and ܤare not similar matrices.
3. Show that ൌ ቂ ቃ Ƭ ܤൌ ቂ ቃ are similar matrices.
െ െ
Solution: (i) ሾ ܣെ ɉ ሿܺ ൌ Ͳ
ͷെɉ ͷ ͳݔ Ͳ
ቂ ቃ ቂ ݔቃ =ቂ ቃ
െʹ Ͳെɉ ʹ Ͳ
ȁ ܣെ ɉ ȁ ൌ Ͳ
ͷെɉ ͷ
ቚ ቚൌͲ
െʹ Ͳെɉ
ɉʹ െ ͷɉ ͳͲ ൌ Ͳ
ฺ ɉ ൌ ʹǤͷǡ ʹǤͷ
(ii)ሾ ܤെ ɉ ሿܺ ൌ Ͳ
ͳെɉ ʹ ͳݔ Ͳ
ቂ ቃ ቂ ݔቃ =ቂ ቃ
െ͵ Ͷ െ ɉ ʹ Ͳ
ȁ ܤെ ɉ ȁ ൌ Ͳ
MATRIX THEORY | 19
ͳെɉ ʹ
ቚ ቚൌͲ
െ͵ Ͷെɉ
ɉʹ െ ͷɉ ͳͲ ൌ Ͳ
ฺ ɉ ൌ ʹǤͷǡ ʹǤͷ
Since, characteristic equation and eigen values of ܤ&ܣare same. ܤ&ܣare similar Matrices.
Theorem: 8. Show that characteristic roots of a Hermitian matrix are all real.
Proof: Let ܣbe the given matrix and let ߣ be the characteristic root of ܣand ܺ be the
characteristic vector of ܣ.
By definition
ܺܣൌ ߣܺ ǥ ǥ ǥ ǥ ሺͳሻ
ሺܺܣሻ כൌ ሺߣܺሻכ
כ ܺ כܣൌ ߣכ ܺ כ
ܣ כ ܺ ൌ ߣ כ ܺ כǥ ǥ ǥ ሾ כܣൌ ܣݏܽܣሿ
On multiplying both sides by ߣܺ ൌ ܺܣ
ܣ כ ܺ ሺߣܺሻ ൌ ߣ כ ܺ כሺܺܣሻ
ฺ ߣሺܺ ܺܣ כሻ ൌ ߣ כሺܺ ܺܣ כሻ
ฺ ߣ ൌ ߣכ
ߣ
Characteristic roots of a Hermitian matrix are all real
Theorem: 9. Show that eigen vectors corresponding to distinct eigen values of a Hermitian
matrix are orthogonal.
Proof: Let ԢܣԢ be the given matrix, which is Hermitian.
Let ߣͳ and ߣʹ are the distinct eigen values. Let ܺͳ and ܺʹ are corresponding eigen vectors.
Then ͳܺܣൌ ߣͳ ܺͳ & ʹܺܣൌ ߣʹ ܺʹ
Now ሺ ͳܺܣሻ כൌ ሺߣͳ ܺͳ ሻכ
כܣ כͳܺ ൌ ߣͳ ܺͳ כǥ ǥ ǥ Ǥ Ǥ ሺߣͳ ݈ܴ݅ܽ݁ݏሻ
ܣ כͳܺ ൌ ߣͳ ܺͳ כሾ כܣൌ ܣݏܽܣሿ
Multiplication with ܺʹ on both sides
ʹܺ כܣ כͳܺ ൌ ߣͳ ܺͳʹܺ כ
ʹܺ ʹߣ כͳܺ ൌ ߣͳ ܺͳ ʹܺ כሾ ʹܺܣ ൌ ߣʹ ܺʹ ሿ
ܺͳ ʹܺ ʹߣ כെ ߣͳ ܺͳ ʹܺ כൌ Ͳ
ܺͳ ʹܺ כሺߣʹ െ ߣͳ ሻ ൌ Ͳ
20 | APPLIED MATHEMATICS - III
SOLVED EXAMPLES
ૡ െૡ െ
1. Show that the matrix ൌ െ െ൩ satisfies the characteristic equation and
െ
hence determine (i) െ (ii) .
Solution: the characteristic equation of matrix ܣis
ͺെߣ
െͺ െʹ
อ Ͷ
െ͵ െ ߣ െʹ อ ൌ Ͳ
͵
െͶ ͳെߣ
͵ ʹ
െߣ ߣ െ ሺെͳͳ ͳͶ ͺሻߣ ሾͺሺെ͵ െ ͺሻ ͺሺͶ ሻ െ ʹሺെͳ ͻሻሿ ൌ Ͳ
െߣ͵ ߣʹ െ ͳͳߣ ൌ Ͳ
ߣ͵ െ ߣʹ ͳͳߣ െ ൌ Ͳ
By Cayley Hamilton theorem
͵ܣെ ʹܣ ͳͳ ܣെ ܫൌ Ͳ ǥ ǥ ǥ ǥ ሺͳሻ
Verification Part
ͺ െͺ െʹ ͺ െͺ െʹ ʹ െ͵ʹ െʹ
Now ʹܣൌ Ͷ െ͵ െʹ൩ Ͷ െ͵ െʹ൩ ൌ ͳͶ െͳͷ െͶ൩
͵ െͶ ͳ ͵ െͶ ͳ ͳͳ െͳ ͵
ͺ െͺ െʹ ʹ െ͵ʹ െʹ
͵ܣൌ ʹܣܣൌ Ͷ െ͵ െʹ൩ ͳͶ െͳͷ െͶ൩
͵ െͶ ͳ ͳͳ െͳ ͵
Ͷ െͳͲͶ ͳͲ
ൌ ͶͲ െͷͳ െʹ൩
͵͵ െͷʹ ͳ͵
͵ܣ െ ʹܣ ͳͳ ܣെ ܫ
Ͷ െͳͲͶ ͳͲ ʹ െ͵ʹ െʹ ͺ െͺ െʹ
ൌ ͶͲ െͷͳ െʹ൩ െ ͳͶ െͳͷ െͶ൩ ͳͳ Ͷ െ͵ െʹ൩
͵͵ െͷʹ ͳ͵ ͳͳ െͳ ͵ ͵ െͶ ͳ
ͳ Ͳ Ͳ Ͳ Ͳ Ͳ
െ Ͳ ͳ Ͳ൩ ൌ Ͳ Ͳ Ͳ൩
Ͳ Ͳ ͳ Ͳ Ͳ Ͳ
Hence Verified
To find ܣെͳ multiply equation (1) with ܣെͳ we get
ܣെͳ ͵ܣെ ܣെͳ ʹܣ ͳͳܣെͳ ܣെ ܣെͳ ܫൌ ܣെͳ Ͳ ൌ Ͳ
ฺ ܣെͳ ൌ ʹܣെ ܣ ͳͳܫ
ʹ െ͵ʹ െʹ Ͷͺ െͶͺ െͳʹ ͳͳ Ͳ Ͳ
ܣെͳ ൌ ͳͶ െͳͷ െͶ൩ െ ʹͶ െͳͺ െͳʹ൩ Ͳ ͳͳ Ͳ൩
ͳͳ െͳ ͵ ͳͺ െʹͶ Ͳ Ͳ ͳͳ
22 | APPLIED MATHEMATICS - III
െͳͳ ͳ ͳͲ
ൌ െͳͲ ͳͶ ͺ ൩
െ ͺ ͺ
ͳ െͳͳ ͳ ͳͲ
ܣെͳ ൌ െͳͲ ͳͶ ͺ ൩
െ ͺ ͺ
From (1), ͵ܣെ ʹܣ ͳͳ ܣെ ܫ
Multiplying by ܣ
ܣͶ െ ͵ܣ ͳͳ ʹܣെ ܣൌ Ͳ
ܣͶ ൌ ͵ܣെ ͳͳ ʹܣ ܣ
But ͵ܣൌ ʹܣെ ͳͳ ܣ ܫ
ܣ Ͷ ൌ ሾ ʹܣെ ͳͳ ܣ ܫሿ െ ͳͳ ʹܣ ܣ
ൌ ͵ ʹܣെ ܣ ͵ ܫെ ͳͳʹܣ
ൌ ʹͷ ʹܣെ Ͳ ܣ ͵ܫ
ʹ െ͵ʹ െʹ ͺ െͺ െʹ ͳ Ͳ Ͳ
ൌ ʹͷ ͳͶ െͳͷ െͶ൩ െ Ͳ Ͷ െ͵ െʹ൩ ͵ Ͳ ͳ Ͳ൩
ͳͳ െͳ ͵ ͵ െͶ ͳ Ͳ Ͳ ͳ
ʹͲ െ͵ʹͲ Ͳ
ܣͶ ൌ ͳͳͲ െͳͷͻ ʹͲ ൩
ͻͷ െͳͷͻ െͳͻ
ࢉ െ࢈
2. Verify cayley Hamilton theorem for the matrix െࢉ ࢇ ൩, hence find െ if it
࢈ െࢇ
exists.
െߣ ܿ െܾ
Solution: the characteristic equation is െܿ െߣ ܽ ൩ ൌ Ͳ
ܾ െܽ െߣ
ߣ Ͳߣ െ ሺܽ ܾ ܿ ʹ ሻߣ ሾെܿሺെܾܽሻ െ ܾሺܽܿሻሿ ൌ Ͳ
͵ ʹ ʹ ʹ
Ͳ െܿ ͵ െ ܾ ʹ ܿ െ ܽʹ ܿ ܾܿ ʹ ܾ ͵ ܽʹ ܾ
ൌ ܿ ܽ ܿ ܾʹ ܿ
͵ ʹ
Ͳ െܾܽ ʹ െ ܽܿ ʹ െ ܽ͵ ൩
െܾܿ ʹ െ ܾ ͵ െ ܽʹ ܾ ܽܿ ʹ ܾܽ ʹ ܽ͵ Ͳ
Ͳ െܿ ͵ െ ܾ ͵ ܿ െ ܽʹ ܿ ܾܿ ʹ ܾ ͵ ܽʹ ܾ
͵ܣ ሺܽʹ ܾ ʹ ܿ ʹ ሻ ܣൌ ܿ ͵ ܽʹ ܿ ܾ ʹ ܿ Ͳ െܾܽ ʹ െ ܽܿ ʹ െ ܽ͵ ൩
െܾܿ ʹ െ ܾ ͵ െ ܽʹ ܾ ܽܿ ʹ ܾܽ ʹ ܽ͵ Ͳ
Ͳ ܿ െܾ
ሺܽʹ ܾ ʹ ܿ ʹ ሻ െܿ Ͳ ܽ൩
ܾ െܽ Ͳ
Ͳ Ͳ Ͳ
ൌ Ͳ Ͳ Ͳ൩. Hence verification.
Ͳ Ͳ Ͳ
࣊ ࣊Ȁ
3. If ൌ ൨, find ܛܗ܋.
࣊Ȁ
Solution: The characteristic equation is ȁ ܣെ ߣܫȁ ൌ Ͳ
ߨ
ߨെߣ
ተ Ͷ ተൌͲ
ߨ
Ͳ െߣ
ʹ
͵ߨ ߨʹ
ߣʹ െ ߣ ൌͲ
ʹ ʹ
ߨ
ߣ ൌ ߨ ߣݎൌ
ʹ
͵ߨ ߨʹ
Now when dividend ߣ is divided by divisor ߣʹ െ ߣ let the quotient be ߶ሺߣሻ and the
ʹ ʹ
remainder by ܽߣ ܾ
݀݊݁݀݅ݒ݅݀ ൌ ሺ ݐ݊݁݅ݐݑݍൈ ݀݅ݎݏ݅ݒሻ ݎ݁݀݊݅ܽ݉݁ݎ
͵ߨ ߨʹ
ߣ ൌ ߶ሺߣሻ ቆߣʹ െ ߣ ቇ ܽߣ ܾ
ʹ ʹ
put ߣ ൌ ߨ ߨ ൌ െͳ ൌ ܽߨ ܾ ǥ ǥ ǥ Ǥ Ǥ ሺܣሻ
ߨ
put ߣ ൌ ߨȀʹ Ͳ ൌ ܽ ቀ ቁ ܾ ǥ ǥ ǥ ǥ Ǥ Ǥ ሺܤሻ
ʹ
ʹ
Solving (A) and (B), ܽ ൌ െ , ܾ ൌ ͳ
ߨ
͵ߨ ߨʹ ʹ
ߣ ൌ ߶ሺߣሻ ቆߣʹ െ ߣ ቇെ ߣͳ
ʹ ʹ ߨ
ʹ
ܣ ൌ െ ܣ ܫ
ߨ
24 | APPLIED MATHEMATICS - III
ߨ
ʹ ߨ ͳ
Ͷ ͳ Ͳ
ܣൌ െ ൦ ߨ ൪ ቂͲ ቃ ൌ െͳ െ ൩
ʹ
ߨ Ͳ ͳ
Ͳ ͳ
ʹ
4. If ൌ ൩ show that for every integer ǡ ൌ െ െ ࡵ Hence, find
ͳെߣ Ͳ Ͳ
Solution: The characteristic equation of A is ͳ Ͳെߣ ͳ ൩ൌͲ
Ͳ ͳ Ͳെߣ
ʹ
ሺͳ െ ߣሻሾߣ െ ͳሿ െ Ͳ Ͳ ൌ Ͳ
ʹߣ െ ͳ െ ߣ͵ ߣ ൌ Ͳ
͵ߣ െ ߣʹ െ ߣ ͳ ൌ Ͳ
By Cayley-Hamiltion theorem this equation is satisfied by A
͵ܣ െ ʹܣെ ܣ ܫൌ Ͳ……………..(1)
͵ܣ ൌ ܣ ʹܣെ ͳ ………………..(2)
We prove the required result by the method of mathematical induction.
Let the result be true for ൌ i.e. suppose ݇ܣൌ ݇ܣെʹ ʹܣെ ܫbe true
Now, multiply the equation by A
݇ܣ ͳ ൌ ݇ܣെͳ ͵ܣെ ܣ
But by (1), ͵ܣെ ܣൌ ʹܣെ ܫ
݇ܣ ͳ ൌ ݇ܣെͳ ʹܣെ ܫ
݇ܣ ͳ ൌ ܣሺ݇ͳሻെʹ ʹܣെ ܫ
Hence, the result is true for ݊ ൌ ݇ ͳǤ
But by (2), the result is true for ݊ ൌ ͵.
Hence, by mathematical induction, it is true for ݊ ൌ Ͷǡͷ ǥ ǥ ݂ ݈݈݊ܽݎ ͵Ǥ
Hence, ݊ܣൌ ݊ܣെʹ ʹܣെ (…………ܫ3)
To find ܣͷͲ ǡwe put successively, ݊ ൌ ʹǡͶ ǥ ǥ ͶǡͶͺǡͷͲ݅݊ሺ͵ሻ
ʹܣ ൌ ܫ ʹܣെ ܫ
ܣͶ ൌ ʹܣ ʹܣെ ܫ
ܣ ൌ ܣͶ ʹܣെ ܫ
………………….
ܣͶ ൌ ܣͶͶ ʹܣെ ܫ
MATRIX THEORY | 25
݂ ሺߣሻ ൌ ߣͺ
ߣ ͺ ൌ ܽߣ ܾ ǥ ǥ ǥ ሺͳሻ
Putߣ ൌ ʹǤʹ͵
ሺʹǤʹ͵ሻͺ ൌ ܽሺʹǤʹ͵ሻ ܾ
ሺʹǤʹ͵ሻͺ ൌ ʹǤʹ͵ܽ ܾ ՜ ሺܽሻ
Putߣ ൌ െʹǤʹ͵
ሺെʹǤʹ͵ሻͺ ൌ ܽሺെʹǤʹ͵ሻ ܾ
ሺെʹǤʹ͵ሻͺ ൌ െʹǤʹ͵ܽ ܾ ՜ ሺܾሻ
ሺܽሻ ሺܾሻ
ʹǤʹ͵ܽ ܾ ൌ ሺʹǤʹ͵ሻͺ
െʹǤʹ͵ܽ ܾ ൌ ሺെʹǤʹ͵ሻͺ
ʹܾ ൌ ሺʹǤʹ͵ሻͺ ሺʹǤʹ͵ሻͺ
ʹܾ ൌ ሺʹǤʹ͵ሻͺ ሺʹǤʹ͵ሻͺ
ܾʹ ൌ ʹሺʹǤʹ͵ሻͺ
ܾ ൌ ሺʹǤʹ͵ሻͺ
ʹǤʹ͵ ൈ ܽ ሺʹǤʹ͵ሻͺ ൌ ሺʹǤʹ͵ሻͺ
ʹǤʹ͵ܽ ൌ ሺʹǤʹ͵ሻͺ െ ሺʹǤʹ͵ሻͺ
ʹǤʹ͵ܽ ൌ Ͳ
ܽൌͲ
From݁ ݊ ݍሺͳሻ
ߣͺ ൌ Ͳ ൈ ߣ ሺʹǤʹ͵ሻͺ
ߣ ͺ ൌ ሺʹǤʹ͵ሻͺ
ܣ ͺ ൌ ሺʹǤʹ͵ሻͺ Ǥ ܫ
ͳ Ͳ
ܣ ͺ ൌ ሺʹǤʹ͵ሻͺ Ǥ ቂ ቃ
Ͳ ͳ
ሺʹǤʹ͵ሻͺ Ͳ
ܣ ͺ ൌ ቈ
Ͳ ሺʹǤʹ͵ሻͺ
7. If ൌ ቂ ቃ, find െ ૢ
Solution: ࢙࢚ Method
The characteristic equation is ȁ ܣെ ߣܫȁ ൌ Ͳ
ʹെߣ ͵
ቚ ቚൌͲ
Ͳ ͳെߣ
28 | APPLIED MATHEMATICS - III
ߣʹ െ ͵ߣ ʹ ൌ Ͳ
By Cayley Hamilton Theorem,
ʹܣെ ͵ ܣ ʹ ܫൌ Ͳ
ʹܣെ ʹ ܣൌ ܣെ ʹ ܫǥ ǥ ǥ ǥ Ǥ ሺͳሻ
premultiplying by ܣ,
͵ܣെ ʹ ʹܣൌ ʹܣെ ʹ ܫܣൌ ʹܣെ ʹ ܣൌ ܣെ ʹ ܫǥ ǥ ǥ ሺʹሻ
premultiplying again by ܣ,
ܣͶ െ ʹ ͵ܣൌ ʹܣെ ʹܫ ʹܣ
͵ܣെ ʹ ʹܣൌ ܣെ ʹܫ
ʹͲ ͳͻ
repeating, ܣ െ ʹܣ ൌ ܣെ ʹܫ
Ͳʹܣ െ ʹ ͻͳܣ ܣൌ ʹ ܣെ ʹܫ
ʹ ͵ ͳ Ͳ ʹ
ൌ ʹቂ ቃ െ ʹቂ ቃൌቂ ቃ
Ͳ ͳ Ͳ ͳ Ͳ Ͳ
ࢊ Method: The characteristic equation is ȁ ܣെ ߣܫȁ ൌ Ͳ
ʹെߣ ͵
ቚ ቚൌͲ
Ͳ ͳെߣ
ߣʹ െ ͵ߣ ʹ ൌ Ͳ
ሺߣ െ ʹሻሺߣ െ ͳሻ ൌ Ͳ
ߣ ൌ ʹ ߣݎൌ ͳ
Now, ߣʹͲ െ ʹߣͳͻ ߣ be the dividend and ߣʹ െ ͵ߣ ʹ be divisor.
݀݊݁݀݅ݒ݅݀ ൌ ሺ ݐ݊݁݅ݐݑݍൈ ݀݅ݎݏ݅ݒሻ ݎ݁݀݊݅ܽ݉݁ݎ
Ͳʹߣ െ ʹߣͳͻ െ ߣ ൌ ߶ሺߣሻሺߣʹ െ ͵ߣ ʹሻ ሺܽߣ ܾሻ
Put ߣ ൌ ʹ
ʹʹͲ െ ʹሺʹͳͻ ሻ ʹ ൌ ߶ሺʹሻሺͲሻ ʹܽ ܾ
ܽʹ ܾ ൌ ʹ ǥ ǥ ǥ ǥ ǥ ሺܣሻ
Put ߣ ൌ ͳ
ͳʹͲ െ ʹሺͳሻͳͻ ͳ ൌ ߶ሺͳሻሺͲሻ ܽ ܾ
ܽ ܾ ൌ Ͳ ǥ ǥ ǥ ǥ ǥ ሺܤሻ
Solving (A) and (B), ܽ ൌ ʹǡ ܾ ൌ െʹ
ߣʹͲ െ ʹߣͳͻ ߣ ൌ ߶ሺߣሻሺߣʹ െ ͵ߣ ʹሻ ʹߣ െ ʹ
Ͳʹܣ െ ʹ ͻͳܣ ܣൌ ߶ሺܣሻሺ ʹܣെ ͵ ܣ ʹܫሻ ʹ ܣെ ʹܫ
but ʹܣെ ͵ ܣ ʹ ܫൌ Ͳ (by Cayley Hamilton theorem)
Ͳʹܣ െ ʹ ͻͳܣ ܣൌ ʹ ܣെ ʹܫ
ʹ ͵ ͳ Ͳ ʹ
ൌ ʹቂ ቃ െ ʹቂ ቃൌቂ ቃ
Ͳ ͳ Ͳ ͳ Ͳ Ͳ
MATRIX THEORY | 29
8. Find the scalar solution of the matrix equation െ െ ࡵ ൌ where
is ൈ matrix.
Solution: By replacing ܣby ߣ we have
ߣ͵ െ ͳͳߣʹ ͵ߣ െ ͵ ൌ Ͳ ǥ ǥ ǥ Ǥ Ǥ ሺͳሻ
ฺ ሺߣ െ ʹሻሺߣ െ ͵ሻሺߣ െ ሻ ൌ Ͳ
ߣ ൌ ʹǡ ͵ǡ are the roots of (1)
͵ܣെ ͳͳ ʹܣ ͵ ܣെ ͵ ܫൌ Ͳ
ฺ ሺ ܣെ ʹܫሻሺ ܣെ ͵ܫሻሺ ܣെ ܫሻ ൌ Ͳ
ܣ െ ʹ ܫൌ Ͳ
ͳ Ͳ ʹ Ͳ
ܣ ൌ ʹቂ ቃൌቂ ቃ
Ͳ ͳ Ͳ ʹ
ฺ ܣെ ͵ ܫൌ Ͳ
ͳ Ͳ ͵ Ͳ
ฺ ܣൌ ͵ ܫൌ ͵ ቂ ቃൌቂ ቃ
Ͳ ͳ Ͳ ͵
ฺ ܣെ ܫൌ Ͳ
ͳ Ͳ Ͳ
ܣ ൌ ܫൌ ቂ ቃൌቂ ቃ
Ͳ ͳ Ͳ
which is the required scalar solution.
9. Find characteristic roots of the matrix ൌ ቂ ቃ also verity Cayley - Hamiton
theorem, also express െ െ ૠ െ െ ࡵ as a linear polynomial in A.
Solution: (a) Eigen values:
(i) ሾ ܣെ ߣ ܫሿ ൈൌ Ͳ
ͳെߣ Ͷ ͳݔ Ͳ
ቂ ቃ ቂ ݔቃ ൌ ቂ ቃ
ʹ ͵െߣ ʹ Ͳ
(ii) ȁ ܣെ ߣܫȁ ൌ Ͳ
ͳെߣ Ͷ
ቚ ቚൌͲ
ʹ ͵െߣ
ߣʹ െ ሺݏݐ݈݈݊݁݉݁݁ܽ݊݃ܽ݅ܦ݂݉ݑݏሻߣ ȁܣȁ ൌ Ͳ
ߣʹ െ ሺͶሻߣ ሺെͷሻ ൌ Ͳ
ߣʹ െ Ͷߣ െ ͷ ൌ Ͳ
ߣ ൌ ͷǡ െͳሾ ሿ
(b) Cayley - Hamilton theorem
ͻ ͳ ͳ Ͷ ͳ Ͳ
Consider ʹܣെ Ͷ ܣെ ͷ ܫൌ ቂ ቃ െ Ͷቂ ቃ െ ͷቂ ቃ
ͺ ͳ ʹ ͵ Ͳ ͳ
30 | APPLIED MATHEMATICS - III
ͻ ͳ Ͷ ͳ ͷ Ͳ Ͳ Ͳ
ൌቂ ቃെቂ ቃെቂ ቃൌቂ ቃ
ͺ ͳ ͺ ͳʹ Ͳ ͷ Ͳ Ͳ
Cayley - Hamilton Theorem is verified
(c) To find െ : ʹܣെ Ͷ ܣെ ͷ ܫൌ Ͳ
ʹܣ Ǥ ܣെͳ െ ͶܣǤ ܣെͳ െ ͷܫǤ ܣെͳ ൌ ͲǤ ܣെͳ
ܣǤ ܣǤ ܣെͳ െ Ͷ ܫെ ͷܣെͳ ൌ Ͳ
ͷܣെͳ ൌ ܫܣെ Ͷܫ
ͳ ͳ
ܣ െͳ ൌ ሾ ܣെ Ͷ ܫሿ ൌ ቀቂͳ Ͷቃ െ Ͷ ቂͳ Ͳ
ቃቁ
ͷ ͷ ʹ ͵ Ͳ ͳ
ͳ ͳ Ͷ Ͷ Ͳ
ൌ ቀቂ ቃെቂ ቃቁ
ͷ ʹ ͵ Ͳ Ͷ
ͳ െ͵ Ͷ
ܣ െͳ ൌ ቂ ቃ
ͷ ʹ െͳ
A3 -2A+3I
ͷ
ͷ െ ͶܣͶ െ ͵ܣ ͳͳ ʹܣെ ܣെ ͳͲ ቀ ܣൌ ܣ ͵ܣൌ ߣ
തതതതതതതതതതതതതതതതതതതതതതതതതതതതതതതതതതതതതതതതതതത
ሺ݀ሻ ʹܣെ Ͷ ܣെ ͷሻܣ ʹ ܣ
െܣͷ െ ͶܣͶ െ ͷ ͵ܣ
െ
െʹ͵ܣ
െʹ ͵ܣ ͳͳ ʹܣെ ܣെ ͳͲ ܫቀ ൌ െʹܣ
ʹܣ
െʹ ͵ܣ ͺ ʹܣ ͳͲܣ
െ െ
ʹ ͵ʹܣ
͵ ܣെ ͳͳ ܣെ ͳͲ ܫቀ ൌ͵
ʹܣ
െ͵ ʹܣെ ͳʹ ܣെ ͳͷܫ
െ
ܣ ͷܫ
ݐ݊݁݀݅ݒ݅ܦ ൌ ݀݅ ݎݏ݅ݒൈ ܳ ݐ݊݁݅ݐݑ ܴ݁݉ܽ݅݊݀݁ݎ
ܣ ͷ െ ͶܣͶ െ ͵ܣ ͳͳ ʹܣെ ܣെ ͳͲܫ
ൌ ሺ ʹܣെ Ͷ ܣെ ͷܫሻ ൈ ሺ ͵ܣെ ʹ ܣ ͵ܫሻ ሺ ܣ ͷܫሻ
ൌ Ͳ ൈ ሺ ʹܣെ ʹ ܣ ͵ܫሻ ሺ ܣ ͷܫሻ
ൌ ܣ ͷܫ
ͳ Ͷ ͳ Ͳ
ൌቂ ቃ ͷቂ ቃ
ʹ ͵ Ͳ ͳ
ͳ Ͷ ͷ Ͳ
ൌቂ ቃቂ ቃ
ʹ ͵ Ͳ ͷ
Ͷ
ൌቂ ቃ
ʹ ͺ
MATRIX THEORY | 31
SOLVED EXAMPLES
ૡ െ
1. Show that the matrix ൌ ૠ െ൩ is digonalisable. Find the transforming
െ
matrix and diagonal matrix.
ͺെߣ െ ʹ
Sol: The characteristic equation is െ െ ߣ െͶ ൩ ൌ Ͳ
ʹ െͶ ͵ െ ߣ
ሺͺ െ ߣሻሾሺ െሻሺ͵ െ ߣሻ െ ͳሿ ሾെሺ͵ െ ߣሻ ͺሿ ʹሾʹͶ െ ʹሺ െ ߣሻሿ ൌ Ͳ
͵ߣ െ ͳͺߣʹ Ͷͷߣ ൌ Ͳ
ሺߣʹ െ ͳͺߣ Ͷͷሻ ൌ Ͳ
ߣ ሺߣ െ ͵ሻሺߣ െ ͳͷ ൌ Ͳ
ߣ ൌ Ͳǡ͵ǡ ͳͷǤ
Since, all Eigen Values are distinct the matrix A is Digonalisable.
ͺ െ ʹ ܺͳ Ͳ
(i) For ߣ ൌ Ͳǡ ሾ ܣെ ߣͳ ܫሿܺ ൌ Ͳ gives െͶ൩ ܺʹ ൩ ൌ Ͳ൩
ʹ െͶ ͵ ܺ͵ Ͳ
ͺͳ െ ʹݔ ʹ ͵ݔൌ Ͳ
െ ͳݔ ʹݔെ Ͷ ͵ݔൌ Ͳ
By crammers’s rule
ͳ െʹ ͵
ൌ ൌ
െ ʹ ͺ ʹ ͺ െ
ቚ ቚ ቚ ቚ ቚ ቚ
െͶ െ െͶ െ
ͳ ʹ ͵ ͳ ʹ ͵
ൌ ൌ ฺ ൌ ൌ
ͳͲ ʹͲ ʹͲ ͳ ʹ ʹ
Corresponding to the Eigen value 0 the Eigen vector is ሾͳǡʹǡ ʹሿԢ
ͷ െ ʹ ܺͳ Ͳ
ሾ ሿ
(ii) For ߣ ൌ ͵ǡ ܣെ ߣ ܺ ܫൌ Ͳ gives െ Ͷ െͶ൩ ܺʹ ൩ ൌ Ͳ൩
ʹ െͶ Ͳ ܺ͵ Ͳ
ͷܺͳ െ ܺʹ ʹܺ͵ ൌ Ͳ
െܺͳ Ͷܺʹ െ Ͷܺ͵ ൌ Ͳ
By crammers’s rule
ͳ െʹ ͵
ൌ ൌ
െ ʹ ͷ ʹ ͷ െ
ቚ ቚ ቚ ቚ ቚ ቚ
Ͷ െͶ െ െͶ െ Ͷ
ܺͳ ܺʹ ܺ͵ ܺͳ ܺʹ ܺ͵
ൌ ൌ ฺ ൌ ൌ
ͳ ͺ െͳ ʹ ͳ െʹ
͵ ǡ ሾʹͳ െ ʹሿԢ
MATRIX THEORY | 33
െ െ ʹ ܺͳ Ͳ
(iii) For ߣ ൌ ͳͷǡ ሾ ܣെ ߣ͵ ܫሿܺ ൌ Ͳ gives െ െͺ െͶ൩ ܺʹ ൩ ൌ Ͳ൩
ʹ െͶ ͳʹ ܺ͵ Ͳ
െܺͳ െ ܺʹ ʹܺ͵ ൌ Ͳ
െܺͳ െ ͺܺʹ െ Ͷܺ͵ ൌ Ͳ
ฺ െܺͳ ܺʹ െ ʹܺ͵ ൌ Ͳ
െܺͳ ͺܺʹ Ͷܺ͵ ൌ Ͳ
By crammers’s rule
ͳ െ ʹ ͵
ൌ ൌ
െʹ െʹ
ቚ ቚ ቚ ቚ ቚ ቚ
ͺ Ͷ Ͷ ͺ
ܺͳ ܺʹ ܺ͵ ܺͳ ܺʹ ܺ͵
ൌ ൌ ฺ ൌ ൌ
ͶͲ ͶͲ ʹͲ ʹ െʹ ͳ
ͳ ͷǡ ሾʹ െ ʹͳሿԢ
ͺ െ ʹ
Since, ܯെͳ ܯܣൌ ܦǡthe given matrix ܣൌ െ െͶ൩ is diagonalised to diagonal matrix
ʹ െͶ ͵
ࡰ ൌ ൩ by transforming matrix ࡹ ൌ െ൩
െ
െ
2. Show that the matrix ൌ ൩ is similar to a diagonal matrix. Also find the
െ
transforming matrix and diagonal matrix.
Solution: Characteristic matrix is
ʹ െʹ ͵ ͳ Ͳ Ͳ ʹ െ ߣ െʹ ͵
ܣെ ߣ ܫൌ ͳ ͳ ͳ ൩ െ ߣ ͲͳͲ൩ ൌ ͳ ͳ െ ߣ ͳ ൩
ͳ ͵ െͳ Ͳ Ͳ ͳ ͳ ͵ െͳ െ ߣ
Characteristic equation is
ሺെͳሻ͵ ߣ͵ ʹߣʹ ͷߣ െ ൌ Ͳ
ฺ ߣ͵ െ ʹߣʹ െ ͷߣ ൌ Ͳ
ฺ ሺߣ െ ͳሻሺߣʹ െ ߣ െ ሻ ൌ Ͳ
ฺ ሺߣ െ ͳሻሺߣ െ ͵ሻሺߣ ʹሻ ൌ Ͳ
ฺ ߣ ൌ ͳǡ െʹǡ ͵
For ߣͳ ൌ ͳǡ ሺ ܣെ ߣͳ ܫሻܺͳ ൌ Ͳ
34 | APPLIED MATHEMATICS - III
ͳ െʹ ͵ ͳݔ Ͳ
ݔ
gives ͳ Ͳ ͳ ൩ ʹ ൩ ൌ Ͳ൩
ͳ ͵ െʹ ͵ݔ Ͳ
ͳ െʹ ͵
Now to find the rank of ͳ Ͳ ͳ ൩
ͳ ͵ െʹ
ͳ െʹ ͵
ܴʹ െ ܴͳ ǡ ܴ͵ െ ܴͳ gives ̱ Ͳ ʹ െʹ൩
Ͳ ͷ െͷ
ͳ െʹ ͵
ͳ ͳ
ܴʹ ǡ ܴ͵ gives ̱ Ͳ ͳ െͳ൩
ʹ ͷ
Ͳ ͳ െͳ
ͳ െʹ ͵
ܴ͵ െ ܴʹ gives ̱ Ͳ ͳ െͳ൩
Ͳ Ͳ Ͳ
equation becomes ͳݔെ ʹ ʹݔ ͵ ͵ݔൌ Ͳ
Ͳ ͳݔ ʹݔെ ͵ݔൌ Ͳ
ͳݔ െʹݔ ͵ݔ
ൌ ൌ
െʹ ͵ ͳ ͵ ͳ െʹ
ቚ ቚ ቚ ቚ ቚ ቚ
ͳ െͳ Ͳ െͳ Ͳ ͳ
ͳݔ െ͵ݔ ʹݔ
ฺ ൌ ൌ
െͳ െͳ ͳ
െͳ
ͳܺ ൌ ͳ ൩
ͳ ߣൌͳ
For ߣʹ ൌ െʹǡ ሺ ܣെ ߣʹ ܫሻܺʹ ൌ Ͳ
Ͷ െʹ ͵ ͳݔ Ͳ
gives ͳ ͵ ͳ൩ ʹݔ൩ ൌ Ͳ൩
ͳ ͵ ͳ ͵ݔ Ͳ
െͳͳ
ʹܺ ൌ െͳ ൩
ͳͶ ߣൌെʹ
For ߣ͵ ൌ ͵ǡ ሺ ܣെ ߣ͵ ܫሻܺ͵ ൌ Ͳ
ͳ െʹ ͵ ͳݔ Ͳ
gives ͳെʹ ͳ ൩ ʹݔ൩ ൌ Ͳ൩
ͳ ͵ െͶ ͵ݔ Ͳ
ͳ
͵ܺ ൌ ͳ൩
ͳ ߣൌ͵
MATRIX THEORY | 35
ܴʹ െ ܴͳ ʹ െͳ െͳ ܺͳ Ͳ
By Ͳ Ͳ Ͳ ൩ ܺʹ ൩ ൌ Ͳ൩
െሺͳΤͶሻ ʹܴͳ
Ͳ Ͳ Ͳ ܺ͵ Ͳ
ͳܺʹ െ ܺʹ െ ܺ͵ ൌ Ͳ
The rank of the coefficient matrix is 1. The number of unknown is 3. Hence, there are
͵ െ ͳ ൌ ʹ linearly independent solutions.
Putting ܺʹ ൌ Ͳܽ݊݀ܺͳ ൌ ͳǡ ͵ܺݐ݁݃݁ݓൌ ʹǤ
Putting ܺ͵ ൌ Ͳܽ݊݀ܺͳ ൌ ͳǡ ʹܺݐ݁݃݁ݓൌ ʹǤ
Corresponding to the Eigen valuesെͳ, we get the following two linearly independent
Eigen vectors.
ܺͳ ൌ ሾͳǡ Ͳǡ ʹሿԢ ܽ݊݀ܺʹ ൌ ሾͳʹͲሿԢ
36 | APPLIED MATHEMATICS - III
െͳʹ Ͷ Ͷ ܺͳ Ͳ
(ii) For ߣ ൌ ͵ǡ ሾ ܣെ ߣʹ ܫሿܺ ൌ Ͳ gives െͺ Ͳ Ͷ൩ ܺʹ ൩ ൌ Ͳ൩
െͳ ͺ Ͷ ܺ͵ Ͳ
ܴʹ െ ܴͳ െͳʹ Ͷ Ͷ ܺͳ Ͳ
By
ܴ͵
ܴͳ Ͷ െͶ Ͳ൩ ܺʹ ൩ ൌ Ͳ൩
െͶ Ͷ Ͳ ܺ͵ Ͳ
െͳʹ Ͷ Ͷ ܺͳ Ͳ
By ܴ͵ ܴʹ Ͷ െͶ ܺ
Ͳ൩ ʹ ൩ ൌ Ͳ൩
Ͳ Ͳ Ͳ ܺ͵ Ͳ
െሺͳΤͶሻ ܴͳ ͵ െͳ െͳ ܺͳ Ͳ
ͳ െͳ Ͳ ൩ ܺ ʹ ൩ ൌ Ͳ൩
ሺͳΤͶሻ ܴʹ
Ͳ Ͳ Ͳ ܺ ͵ Ͳ
ͳܺ͵ െ ܺʹ െ ܺ͵ ൌ Ͳܽ݊݀ܺͳ െ ܺʹ ൌ Ͳ
ͳܺ ൌ ܺʹ
Putting ܺʹ ൌ ͳǡ we get ܺͳ ൌ ͳ
͵ܺ ൌ ʹǤ
Corresponding to Eigen value 3, we get the following Eigen vector.ܺ ൌ ሾͳͳʹሿ
Although Eigen values of A are not distinct the geometric multiplicity of each Eigen
value is equal to it’s algebraic multiplicity, A is diagnosable.
െͻ Ͷ Ͷ
Since, ܯെͳ ܯܣൌ ܦǡ the matrix ܣൌ െͺ ͵ Ͷ൩will be digonalised to the diagonal matrix
ͳ ͺ
െ
D = െ ൩ by transforming matrix ࡹ ൌ ൩
െ െ
4. Show that the matrix ൌ ൩ is similar to a diagonal matrix. Find also the
െ െ
transforming matrix , the diagonal matrix.
Solution: Characteristic equation is
ͳ െ െͶ ͳ Ͳ Ͳ ͳ െ ߣ െ െͶ
ܣെ ߣ ܫൌ Ͳ Ͷ ʹ ൩ െ ߣ ͲͳͲ൩ ൌ Ͳ Ͷ െ ߣ ʹ ൩
Ͳ െ െ͵ Ͳ Ͳ ͳ Ͳ െ െ͵ െ ߣ
Characteristic equation is
ሺെͳሻ͵ ߣ͵ ʹߣʹ െ ሺͳሻߣ ൌ Ͳ
ฺ െߣ͵ ʹߣʹ െ ߣ ൌ Ͳ
ฺ ߣ͵ െ ʹߣʹ ߣ ൌ Ͳ
MATRIX THEORY | 37
ฺ ߣሺߣ െ ʹߣ ͳሻ ൌ Ͳ
ฺ ߣ ൌ Ͳǡ ߣ ൌ ͳǡ ͳ
For ߣ ൌ Ͳǡ ሺ ܣെ ߣܫሻܺ ൌ Ͳ
ͳ െ െͶ ͳݔ Ͳ
gives Ͳ Ͷ ʹ ൩ ʹݔ൩ ൌ Ͳ൩
Ͳ െ െ͵ ͵ݔ Ͳ
ͳ െ െͶ
To find the rank of Ͳ Ͷ ʹ ൩
Ͳ െ െ͵
ͳ െ െͶ
ͳ ͳ
ܴʹ ǡ െ ܴ͵ gives ̱ Ͳ ʹ ͳ ൩
ʹ ͵
Ͳ ʹ ͳ
ͳ െ െͶ
ܴ͵ െ ܴʹ gives ̱ Ͳ ʹ ͳ ൩
Ͳ Ͳ Ͳ
ͳݔ െ ʹݔെ Ͷ ͵ݔൌ Ͳ
Ͳ ͳݔ ʹ ʹݔ ͵ݔൌ Ͳ
By Cramer’s Rule,
ͳݔ െʹݔ ͵ݔ
ൌ ൌ
െ െͶ ͳ െͶ ͳ െ
ቚ ቚ ቚ ቚ ቚ ቚ
ʹ ͳ Ͳ ͳ Ͳ ʹ
ͳݔെ͵ݔ ʹݔ ʹ
ൌ ൌ ͳܺ ൌ െͳ൩
ʹ ͳ ʹ
ʹ
For ߣʹ andߣ͵ ൌ ͳǡ ͳ
Algebraic multiplicity is ʹ to find rank of matrix ܣെ ߣܫ.
We have ሺ ܣെ ߣܫሻܺ ൌ Ͳ
Ͳ െ െͶ ͳݔ Ͳ
Ͳ ͵ ʹ ൩ ʹݔ൩ ൌ Ͳ൩
Ͳ െ െͶ ͵ݔ Ͳ
Ͳ െ െͶ
Rank of Ͳ ͵ ʹ ൩
Ͳ െ െͶ
Ͳ െ െͶ
ܴ͵ ܴʹ gives Ͳ ͵ ʹ ൩
Ͳ Ͳ Ͳ
Ͳ െ െͶ
ͳ
ܴʹ ܴͳ gives Ͳ Ͳ Ͳ ൩
ʹ
Ͳ Ͳ Ͳ
38 | APPLIED MATHEMATICS - III
Ͳ െ െͶ ͳݔ Ͳ
ݔ
Ͳ Ͳ Ͳ ൩ ʹ ൩ ൌ Ͳ൩
Ͳ Ͳ Ͳ ͵ݔ Ͳ
ͳݔͲ െ ʹݔെ Ͷ ͵ݔൌ Ͳ
ʹݔൌ ʹǡ ͳݔൌ Ͳ
െ ʹݔൌ െͶ ͵ݔǡ ͵ݔൌ ͵
Ͳ
ʹܺ ൌ ʹ൩
͵
Solving, ͳݔൌ ͳǡ ʹݔൌ Ͳǡ ͵ݔൌ Ͳ
ͳ
͵ܺ ൌ Ͳ൩
Ͳ ߣൌͳ
Thus Matrix ܮis given by
ʹ Ͳ ͳ
ܮൌ ሾܺͳ ܺʹ ܺ͵ ሿ ൌ െͳ ʹ Ͳ൩
ʹ െ͵ Ͳ
Ͳ ͵ ʹ
ܮ െͳ ൌ Ͳ ʹ ͳ ൩
ͳ െ െͶ
Ͳ Ͳ Ͳ
ܮെͳ ܮܣൌ ܦൌ ͲͳͲ൩
Ͳ Ͳ ͳ
(b). Congruent transformation
If ܣis symmetric matrix or Hermitian matrix and if it can be reduced by congruent
transformation to a diagonal matrix ܦthen ܣand ܦare congruent.
i.e. ܲԢ ܲܣൌ ܣ[ ܦis symmetric matrix]
ܲതԢ ܲܣൌ ܣ[ ܦis hermitian matrix]
SOLVED EXAMPLES
1. Reduce the matrix by congruent transformation to diagonal form where
െ
ൌ െ െ൩
െ
Solution: ܣൌ ͵ܫܣ ͵ܫ
ͳͲ െ͵ ͵ ͳ Ͳ Ͳ ͳ Ͳ Ͳ
െ͵ ͳ െͳ൩ ൌ ͲͳͲ൩ ܣͲͳͲ൩
͵ െͳ ͳ Ͳ Ͳ ͳ Ͳ Ͳ ͳ
MATRIX THEORY | 39
͵ ͵
ܴʹ ܴͳ ǡ ܴ͵ െ ܴͳ
ͳͲ ͳͲ
ͳͲ െ͵ ͵ ͳ Ͳ Ͳ ͳ Ͳ Ͳ
Ͳ ͳȀͳͲ െͳȀͳͲ ൩ൌ ͵ȀͳͲ ͳͲ൩ ܣͲͳͲ൩
Ͳ െͳȀͳͲ ͳȀͳͲ െ͵ȀͳͲ Ͳ ͳ Ͳ Ͳ ͳ
͵ ͵
ʹܥ ܥǡ ͵ܥെ ܥ
ͳͲ ͳ ͳͲ ͳ
ͳͲ Ͳ Ͳ ͳ Ͳ Ͳ ͳ ͵ȀͳͲ െ͵ȀͳͲ
Ͳ ͳȀͳͲ െͳȀͳͲ൩ ൌ ͵ȀͳͲ ͳͲ൩ ܣͲ ͳ Ͳ ൩
Ͳ െͳȀͳͲ ͳȀͳͲ െ͵ȀͳͲ Ͳ ͳ Ͳ Ͳ ͳ
ܴʹ ܴ͵
ͳͲ Ͳ Ͳ ͳ Ͳ Ͳ ͳ ͵ȀͳͲ െ͵ȀͳͲ
Ͳ ͳȀͳͲെͳȀͳͲ൩ ൌ ͵ȀͳͲͳͲ൩ ܣͲ ͳ Ͳ ൩
Ͳ Ͳ Ͳ Ͳ ͳ ͳ Ͳ Ͳ ͳ
ʹܥ ͵ܥ
ͳͲ Ͳ Ͳ ͳ Ͳ Ͳ ͳ ͵ȀͳͲ Ͳ
Ͳ ͳȀͳͲͲ൩ ൌ ͵ȀͳͲͳͲ൩ ܣͲ ͳ Ͳ൩
Ͳ Ͳ Ͳ Ͳ ͳ ͳ Ͳ Ͳ ͳ
2. Find non singular matrix ܲ such that the matrix ܣis reduced to diagonal form where
͵ െͳ ͳ
ܣൌ െͳ ͵ െͳ൩
ͳ െͳ ͵
By using (i) Similarity transformation ሺܲെͳ ܲܣൌ ܦሻ
(ii) Congruent transformation ܲԢ ܲܣൌ ܦ
Solution: (i) ߣ͵ െ ͻߣʹ ʹͶߣ െ ʹͲ ൌ Ͳǡ ߣ ൌ ͷǡ ʹǡ ʹ
ͳȀξ͵ Ͳ ʹȀξ ͷ Ͳ Ͳ
ܲ ൌ ൦െͳȀξ͵ͳȀξʹ ͳȀξ ൪ ܦൌ ͲʹͲ൩
ͳȀξ͵ ͳȀξʹ െͳȀξ Ͳ Ͳ ʹ
ͳ ͳȀ͵ െͳȀͶ
(ii) Ͳ ͳ ͳȀͶ ൩
Ͳ Ͳ ͳ
Ͳ ͳȀ͵ െͳȀͶ ͵ Ͳ Ͳ
ܲ ൌ Ͳ ͳ ͳȀͶ ൩ ܦൌ ͲͺȀ͵ Ͳ ൩
Ͳ Ͳ ͳ Ͳ Ͳ ͷȀʹ
ഥ ࡼ reduce the hermitian matrix
3. Using congruent transformation ࡼ Ԣ
െ
ൌ ൩ to a diagonal form.
െ െ
40 | APPLIED MATHEMATICS - III
െ െ
2. If ൌ െ ൩ . Find the minimal polynomial of and hence show that the
െ െ
matrix is derogatory.
ͷ െ ߣ െ െ
Solution: Method I: The characteristic equation of ܣis อ െͳ Ͷ െ ߣ ʹ อ ൌ Ͳ
͵ െ െͶ െ ߣ
ሺͷ െ ߣሻሾെሺͳ െ ߣʹ ሻ ͳʹሿ ሾͶ ߣ െ ሿ െ ሾ െ ͵ሺͶ െ ߣሻሿ ൌ Ͳ
ሺͷ െ ߣሻሾെͶ ߣʹ ሿ ሾെʹ ߣሿ െ ሾെ ͵ߣሿ ൌ Ͳ
͵ߣ െ ͷߣʹ ͺߣ െ Ͷ ൌ ͲǢ ߣ͵ െ ʹߣʹ െ ͵ߣʹ ߣ ʹߣ െ Ͷ ൌ Ͳ
ሺߣ െ ʹሻሺߣʹ െ ͵ߣ ʹሻ ൌ Ͳ ሺߣ െ ʹሻሺߣ െ ʹሻሺߣ െ ͳሻ ൌ Ͳ
Hence, the roots of ȁ ܣെ ߣܫȁ ൌ Ͳ are ʹǡ ʹǡ ͳ.
Let us now find the minimal polynomial of ܣ. We know that each characteristic root of ܣis also
a root of the minimal polynomial of ܣ. So if ݂ሺݔሻ is the minimal polynomial of ܣthen ݔെ ͳ
and ݔെ ʹ are the factors of ݂ሺݔሻ.
Let us see whether the polynomial ሺ ݔെ ʹሻሺ ݔെ ͳሻ ൌ ʹ ݔെ ͵ ݔ ʹ annihilates ܣ.
ͷ െ െ ʹ ͳ͵ െͳͺ െͳͺ
ʹ
Now, ܣൌ െͳ Ͷ ʹ ൩ ൌ െ͵ ͳͲ ൩
͵ െ െͶ ͻ െͳͺ െͳͶ
ͳ͵ െͳͺ െͳͺ ͷ െ െ ʹ Ͳ Ͳ Ͳ Ͳ Ͳ
ʹܣ െ ͵ ܣ ʹ ܫൌ െ͵ ͳͲ ൩ െ ͵ െͳ Ͷ ʹ ൩ ͲʹͲ൩ ൌ ͲͲͲ൩
ͻ െͳͺ െͳͶ ͵ െ െͶ Ͳ Ͳ ʹ Ͳ Ͳ Ͳ
݂ ሺݔሻ ൌ ʹ ݔെ ͵ ݔ ʹ annihilates ܣ.
Thus, ݂ሺݔሻ is the monic polynomial of lowest degree that annihilates ܣ. Hence, ݂ሺݔሻ is the
minimal polynomial of ܣ. Since its degree is less than the order of ܣ, ܣis derogatory.
Method II: Characteristic equation of matrix ܣis ȁ ܣെ ߣܫȁ ൌ Ͳ
ͷ െ ߣ െ െʹ
อ െͳ Ͷ െ ߣ ʹ อ ൌ Ͳ
͵ െ െͶ െ ߣ
i.e. ሺߣሻ ൌ Ͳ
ሺߣሻ ൌ ߣ͵ െ ͷߣʹ ͺߣ െ Ͷ
ൌ ሺߣ െ ͳሻሺߣ െ ʹሻሺߣ െ ʹሻ ǥ ǥ ǥ Ǥ ሺͳሻ
Consider the first degree monic polynomial corresponding to each factor of (1)
Thus, ݃ͳ ሺߣሻ ൌ ߣ െ ͳ
݃ͳ ሺܣሻ ൌ ܣെ ܫ
ͷ െ െ ͳ Ͳ Ͳ
ൌ െͳ Ͷ ʹ ൩ െ ͲͳͲ൩ ് Ͳ
͵ െ െͶ Ͳ Ͳ ͳ
MATRIX THEORY | 43
݃ʹ ሺߣሻ ൌ ߣ െ ʹ
݃ʹ ሺܣሻ ൌ ܣെ ʹͲ ് ܫ
ͳ݃ ሺߣሻ and ݃ʹ ሺߣሻ are not the minimal polynomials of ܣas they do not annihilate ܣ.
Now consider second degree monic polynomials formed by factors of (1) taken two at a time.
ͳ݂ ሺߣሻ ൌ ሺߣ െ ͳሻሺߣ െ ʹሻ
݂ʹ ሺߣሻ ൌ ሺߣ െ ʹሻሺߣ െ ʹሻ
݂ͳ ሺߣሻ ൌ ߣʹ െ ͵ߣ ʹ
݂ͳ ሺܣሻ ൌ ʹܣെ ͵ ܣ ʹܫ
ͳ͵ െͳͺ െͳͺ ͷ െ െ ͳ Ͳ Ͳ
ൌ െ͵ ͳͲ ൩ െ ͵ െͳ Ͷ ʹ ൩ ʹ ͲͳͲ൩
ͻ െͳͺ െͳͶ ͵ െ െͶ Ͳ Ͳ ͳ
ൌͲ
ͳ݂ ሺߣሻ annihilates ܣ.
Therefore ݂ͳ ሺߣሻ is a monic polynomial of lower degree which annihilates ܣ.
Hence ݂ͳ ሺߣሻ ൌ ݉ሺߣሻ is the minimal polynomial of ܣ.
Since the degree of ݉ሺߣሻ is ʹ is less than the degree of the characteristic polynomial ሺߣሻ.
Therefore matrix ܣis derogatory.
3. Show that ൌ ൩is non-derogatory.
ͳെߣ ʹ ͵
Sol: The characteristic equation of A is ʹ ͵െߣ Ͷ ൩ൌͲ
͵ Ͷ ͷെߣ
ሺͳ െ ߣሻሾሺ͵ െ ߣሻሺͷ െ ߣሻ െ ͳሿ െ ʹሾʹሺͷ െ ߣሻ െ ͳʹሿ ͵ሾͺ െ ͵ሺ͵ െ ߣሻሿ ൌ Ͳ
ሺͳ െ ߣሻሾെͳ െ ͺߣ ߣʹ ሿ െ ʹሾെʹ െ ʹߣሿ ͵ሾെͳ ͵ߣሿ ൌ Ͳ
͵ߣ െ ͻߣʹ െ ߣ ൌ Ͳ
ߣ ሺߣʹ െ ͻߣ െ ሻ ൌ Ͳ
Since all roots are distinct and since the characteristic equation is satisfied by A.A is non-
derogatory.
4. If the matrix ൌ ൩ find the minimal polynomial and state whether matrix
െૠ െ
is derogatory or non derogatory.
Solution: Method I
The characteristic equation is ȁ ܣെ ߣܫȁ ൌ Ͳ
44 | APPLIED MATHEMATICS - III
ʹെߣ ʹ Ͳ
อ ʹ ͳ െ ߣ ͳ อ ൌ Ͳ
െ ʹ െ͵ െ ߣ
െߣ͵ ͳ͵ߣ െ ͳʹ ൌ Ͳ
ฺ ߣ ൌ ͳǡ ͵ǡ െͶ
All the characteristic roots are distinct.
Minimal polynomial will contain all characteristic roots.
Minimal polynomial ݂ሺߣሻ ൌ ሺߣ െ ͳሻሺߣ െ ͵ሻሺߣ Ͷሻ
Degree of minimal polynomial is equal to order of matrix ܣ.
Given matrix is non-derogatory.
Method II: The characteristic equation is ȁ ܣെ ߣܫȁ ൌ Ͳ
ʹെߣ ʹ Ͳ
อ ʹ ͳ െ ߣ ͳ อ ൌ Ͳ
െ ʹ െ͵ െ ߣ
ฺ െߣ͵ ͳ͵ߣ െ ͳʹ ൌ Ͳ
ฺ ߣ͵ െ ͳ͵ߣ ͳʹ ൌ Ͳ
ሺߣሻ ൌ ሺߣ െ ͳሻሺߣ Ͷሻሺߣ െ ͵ሻ ǥ ǥ ǥ ǥ ǥ Ǥ Ǥ ሺͳሻ
Consider the first degree monic polynomials corresponding to each factor of (1)
Thus ݃ͳ ሺߣሻ ൌ ߣ െ ͳ
݃ͳ ሺߣሻ ൌ ߣ െ ͳ
݃ͳ ሺܣሻ ൌ ܣെ Ͳ ് ܫ
݃ʹ ሺߣሻ ൌ ߣ Ͷ
݃ʹ ሺܣሻ ൌ ܣ ͶͲ ് ܫ
݃͵ ሺߣሻ ൌ ߣ െ ͵
݃͵ ሺܣሻ ് Ͳ
ͳ݃ ሺߣሻǡ ݃ʹ ሺߣሻǡ ݃͵ ሺߣሻ are not the minimal polynomials of ܣas they do not annihilate ܣ.
Now consider the second degree monic polynomials formed by factors of (1) taken two at a
time.
݂ͳ ሺߣሻ ൌ ሺߣ െ ͳሻሺߣ Ͷሻ
݂ʹ ሺߣሻ ൌ ሺߣ Ͷሻሺߣ െ ͵ሻ
݂͵ ሺߣሻ ൌ ሺߣ െ ͳሻሺߣ െ ͵ሻ
݂ͳ ሺߣሻ ൌ ߣʹ ͵ߣ െ Ͷ
݂ͳ ሺܣሻ ൌ ʹܣ ͵ ܣെ ͶͲ ് ܫ
MATRIX THEORY | 45
݂ʹ ሺߣሻ ൌ ߣʹ ߣ െ ͳʹ
ʹ݂ ሺܣሻ ൌ ʹܣ ܣെ ͳʹͲ ് ܫ
݂͵ ሺߣሻ ൌ ߣʹ െ Ͷߣ ͵
͵݂ ሺܣሻ ൌ ʹܣെ Ͷ ܣ ͵Ͳ ് ܫ
ͳ݃ ሺߣሻǡ ݃ʹ ሺߣሻǡ ݃͵ ሺߣሻ. So also ݂ͳ ሺߣሻǡ ݂ʹ ሺߣሻǡ ݂͵ ሺߣሻ do not annihilate matrix ܣ.
Therefore ሺߣሻ is itself the minimal polynomial of degree ͵.
Since the degree of the characteristic polynomial ሺߣሻ is equal to the degree of the minimal
polynomial. Therefore the matrix is non derogatory.
െ
5. Show that the matrix ൌ ൩ is non-derogatory.
െ
Solution: The characteristic equation of ܣis
ʹ െ ߣ െʹ ͵
อ ͳ ͳ െ ߣ ͳ อ ൌ Ͳ
ͳ ͵ െͳ െ ߣ
ሺʹ െ ߣሻሾെሺͳ െ ߣሻሺͳ ߣሻ െ ͵ሿ ʹሾെሺͳ ߣሻ െ ͳሿ ͵ሾ͵ െ ሺͳ െ ߣሻሿ ൌ Ͳ
ሺʹ െ ߣሻሺെͶ ߣʹ ሻ െ ʹሺʹ ߣሻ ͵ሺʹ ߣሻ ൌ Ͳ
͵ߣ െ ʹߣʹ െ ͷߣ ൌ Ͳ
͵ߣ െ ߣʹ െ ߣʹ ߣ െ ߣ ൌ Ͳ
ሺߣ െ ͳሻሺߣʹ െ ߣ െ ሻ ൌ Ͳ
ሺߣ െ ͳሻሺߣ െ ͵ሻሺߣ ʹሻ ൌ Ͳ
ߣ ൌ ͳǡ െʹǡ ͵
Since, all the roots of the characteristic equation are distinct,
݂ሺߣሻ ൌ ሺߣ െ ͳሻሺߣ ʹሻሺߣ െ ͵ሻ
is the minimal polynomial. Hence, the matrix is non-derogatory.
6. If ൌ ቂ ቃǡfind
ʹെߣ ͳ
Solution: The characteristic equation of A is ቚ ቚൌͲ
ͳ ʹെߣ
ሺʹ െ ߣሻʹ െ ͳ ൌ Ͳ
Ͷ െ Ͷߣ ߣʹ െ ͳ ൌ Ͳ
ʹߣ െ Ͷߣ ͵ ൌ Ͳ
ሺߣ െ ͵ሻሺߣ െ ͳሻ ൌ Ͳ
ߣ ൌ ͳǡ͵
46 | APPLIED MATHEMATICS - III
ͳ ͳ ܺͳ Ͳ
(i) For ߣ ൌ ͳǡ ሾ ܣെ ߣܫሿܺ ൌ Ͳ gives ቂ ቃ ൨ ൌ ቂ ቃ
ͳ ͳ ܺʹ Ͳ
ܴʹ ቂͳ ͳ ܺͳ Ͳ
By ܴͳ െ ቃ ൨ ൌ ቂ ቃ
Ͳ Ͳ ܺʹ Ͳ
ͳܺ ܺʹ ൌ Ͳ
Putting ܺʹ ൌ െͳǡ ܺͳ ൌ ͳ
Hence, corresponding to ߣ ൌ ͳ the Eigen vector is ሾͳǡ െͳሿԢ
െͳ ͳ ܺͳ Ͳ
For ߣ ൌ ͵ǡ ሾ ܣെ ߣܫሿܺ ൌ Ͳ givesቂ ቃ ൨ ൌ ቂ ቃ
ͳ െͳ ܺʹ Ͳ
ܺ
By ܴʹ ܴͳ ቂെͳ ͳቃ ͳ ൨ ൌ ቂͲቃ
Ͳ Ͳ ܺʹ Ͳ
െܺͳ ܺʹ ൌ Ͳܽ݊݀ܺͳ ൌ ܺʹ
Putting ܺʹ ൌ ͳǡ ܺͳ ൌ ͳ
Hence, corresponding to ߣ ൌ ͵ǡthe Eigen vector is ሾͳǡͳሿԢ
ͳ ͳ
Modal matrix ܺ ൌ ቂ ቃ
െͳ ͳ
Now ȁܯȁ ൌ ʹ
ͳ ͳ െͳ
ܯ െͳ ൌ ቂ ቃ
ʹ ͳ ͳ
ͳ Ͳ
Now ܦൌ ቂ ቃ
Ͳ ͵
ͷͲ
ܦ ͷͲ ൌ ͳ Ͳ ൨
Ͳ ͵ͷͲ
െͳΤʹ ͳΤʹ ܺͳ Ͳ
(ii) For ߣ ൌ ʹሾ ܣെ ߣܫሿܺ ൌ Ͳ gives ൨ ൨ ൌ ቂ ቃ
ͳΤʹ െͳΤʹ ܺʹ Ͳ
ʹܴ െͳ ͳ ܺͳ Ͳ
By ͳ ቂ ቃ ൨ ൌ ቂ ቃ
ʹܴʹ ͳ െͳ ܺʹ Ͳ
െͳ ͳ ܺ Ͳ
By ܴʹ ܴͳ ቂ ቃ ͳ ൨ ൌ ቂ ቃ
Ͳ Ͳ ܺʹ Ͳ
െܺͳ ܺʹ ൌ Ͳ
ͳܺ ൌ ܺʹ
If ܺʹ ൌ ͳǡ ܺͳ ൌ ͳ. Hence, the Eigen vector is ሾͳǡͳሿԢ
ͳ ͳ
ܯൌቂ ቃ
െͳ ͳ
ȁܯȁ ൌ ʹ
݆ܽ݀ ͳ ͳ ܯെͳ
ܯെͳ ൌ ൌ ቂ ቃ
ȁܯȁ ʹ ͳ ͳ
ͳ Ͳ
Now ܦൌ ቂ ቃ
Ͳ ʹ
ͳ
If ሺܣሻ ൌ ݁ ܣǡ ݂ሺܦሻ݁ ܦൌ ݁ Ͳ൨
Ͳ ݁ʹ
ͳ
If ሺሻ ൌ Ͷ ܣǡ ݂ሺܦሻͶ ܦൌ Ͷ Ͳ൨
Ͳ Ͷʹ
ͳ ͳ ݁ Ͳ ͳ ͳ െͳ
ܣ ݁ ൌ ܯሺሻܯെͳ ൌ ቂ ቃቂ ቃ ቂ ቃ
െͳ ͳ Ͳ ݁ ʹ ʹ ͳ ͳ
ͳ ʹ ͳ െͳ
ൌ ݁ ݁ʹ൨ ቂ ቃ
ʹ െ݁ ݁ ͳ ͳ
ͳ ʹ ʹ
ܣ ݁ ൌ ݁ ݁ ʹ െ݁ ݁ʹ ൨
ʹ െ݁ ݁ ݁݁
Similarly replacing e by 4 we get,
ൌቂ ቃൌቂ ቃ
࢟ ࢟ ࢟ ܐܖܑܛ ࢟ ܐܛܗ܋
9. If ൌ ቂ࢟࢟ቃ prove ࢋ ൌ ࢋ࢟ ൨.
࢟ ܐܛܗ܋ ࢟ ܐܖܑܛ
Solution: The characteristic equation is ȁ ܣെ ߣܫȁ ൌ Ͳ
ݕെߣ ݕ
ฬ ฬൌͲ
ݕ ݕെߣ
ߣʹ െ ʹߣ ݕൌ Ͳ
ߣ ൌ Ͳǡ ߣ ൌ ʹݕ
Now when dividend ൌ ݁ ߣ is divided by divisor ൌ ߣʹ െ ʹߣݕ
MATRIX THEORY | 49
Putting ܺͳ ൌ ͳǡ ʹܺݐ݁݃݁ݓൌ ͳ
The Eigen vector is ሾͳǡ െͳሿԢ
ʹ Ͷ ܺͳ Ͳ
(ii) For ߣ ൌ െ͵ǡ ሾ ܣെ ߣ ܫሿܺ ൌ Ͳ givesቂ ቃ ൨ ൌ ቂ ቃ
ʹ Ͷ ܺʹ Ͳ
ʹ Ͷ ܺͳ Ͳ
Byܴʹ െ ܴͳ ቂ ቃ ൨ ൌ ቂ ቃ
Ͳ Ͳ ܺʹ Ͳ
ͳܺʹ Ͷܺʹ ൌ Ͳ
ฺ ܺͳ ʹܺʹ ൌ Ͳ
Putting ܺʹ ൌ െͳǡ ܺͳ ൌ ʹ
The Eigen vector is ሾʹǡ െͳሿ
ͳ ʹ
ܯൌቂ ቃ ܽ݊݀ȁܯȁ ൌ െ͵
ͳ െͳ
݆ܽ݀Ǥ ܯ ͳ െͳ െʹ
ܯെͳ ൌ ൌെ ቂ ቃ
ȁܯȁ ͵ െͳ ͳ
͵ Ͳ
Now ܦൌ ቂ ቃ
Ͳ െ͵
݂ ሺܣሻ ൌ ܣǤ
͵ Ͳ ͵ Ͳ
ሺሻ ൌ ൨ൌቂ ቃ
Ͳ ሺെ͵ሻ Ͳ െ ͵
ܣ ൌ ݂ܯሺܦሻܯെͳ
ͳ ʹ ͵ Ͳ ͳ െͳ െʹ
ൌቂ ቃቂ ቃ ൬െ ൰ ቂ ቃ
ͳ െͳ Ͳ െ ͵ ͵ െͳ ͳ
ͳ ͵ െʹ ͵ െͳ െʹ
ൌെ ቂ ቃቂ ቃ
͵ ͵ െ ͵ െͳ ͳ
ͳ ͵ െ Ͷ ͵
ൌെ ቂ ቃ
͵ െʹ ͵ െ ͵
െ ͵ Ͷ ͵
ܣ ͵ ൌ ቂ ቃ
ʹ ͵ ͵
െͳ Ͷ
ൌ ͵ ቂ ቃ ൌ ͵ܣ
ʹ ͳ
ൌ ܖ܉ܜ.
ࢉ࢙࢚ ࢙࢚
11. If A =ቂ ቃ ࢚ࢋܜ܉ܐܜ܍ܞܗܚܘൌ ቂ ቃ
െ െ࢙࢚ ࢉ࢙࢚
Solution: (a) Eigen values: (i) ሾ ܣെ ɉ ܫሿ ൈൌ Ͳ
Ͳ െ ɉ ͳ ͳݔ Ͳ
ቂ ቃ ቂ ݔቃ ൌ ቂ ቃ
െͳ Ͳ െ ɉ ʹ Ͳ
MATRIX THEORY | 51
Ǥ Ǥ
12. If A = ቂ ቃ ࢙ࢉ܌ܖܑܖ܍ܐܜെ
Ǥ Ǥ
Solution: Eigen values :
(i) ሾ ܣെ ɉ ሿ ൈൌ Ͳ
ͲǤͷ െ ɉ ͲǤͷ ͳݔ Ͳ
ቂ ቃቂ ݔቃ ൌ ቂ ቃ
ͲǤͷ ͲǤͷ െ ɉ ʹ Ͳ
(ii) ȁ ܣൌ ɉ ȁ ൌ Ͳ
ͲǤͷ െ ɉ ͲǤͷ
ቚ ቚ = 0
ͲǤͷ ͲǤͷ െ ɉ
ɉʹ െ ሺሻɉ ȁȁ ൌ Ͳ
ฺ ɉʹ െ ሺͳሻɉ Ͳ ൌ Ͳ
ฺ ɉʹ െ ɉ ൌ Ͳ
ɉ ൌ Ͳǡͳ
݂ ሺɉሻ ൌ ܽɉ
݂ ሺܣሻ ൌ ܿ ݏെͳ ܣ
݂ ሺɉሻ ൌ ܿ ݏെͳ ɉ
ฺ ܿ ݏെͳ ɉ ൌ ɉ ՜ ሺ ሻ
Put ɉ ൌ Ͳ
ฺ ܿ ݏെͳ ሺͲሻ ൌ ܽ ൈ Ͳ ܾ
ߨ
ฺ ൌ ܾ
ʹ
Put ɉ ൌ ͳ
ฺ ܿݏെͳ ሺͳሻ ൌ ܽ ൈ ͳ ܾ
ฺͲൌܾܽ
ߨ
ฺܽ ൌͲ
ʹ
െߨ
ฺ ܽ ൌ
ʹ
݊ ሺܫሻ
ݍ݁
െɎ Ɏ
ܿ ݏെͳ ɉ ൌ ɉ
ʹ ʹ
Put ɉ = A
െߨ Ɏ
ܿ ݏെͳ ܣൌ
ʹ ʹ
െߨ ߨ
ܿ ݏെͳ ܣൌ ܣ ൈ ܫ
ʹ ʹ
MATRIX THEORY | 53
െߨ ͲǤͷ ͲǤͷ ߨ ͳ Ͳ
ܿ ݏെͳ ܣൌ ൈ ቂ ቃ ቂ ቃ
ʹ ͲǤͷ ͲǤͷ ʹ Ͳ ͳ
ߨ െͲǤͷ െͲǤͷ ͳ Ͳ
ൌ ቄቂ ቃ ቂ ቃቅ
ʹ െͲǤͷ െͲǤͷ Ͳ ͳ
ߨ ͲǤͷ െͲǤͷ
ܿ ݏെͳ ܣൌ ቂ ቃ
ʹ െͲǤͷ ͲǤͷ
ߨ ߨ
ሺߨ െ ߣሻ ቀ െ ߣቁ ൌ Ͳ ߣ ൌ ǡ ߨ
ʹ ʹ
Let (A) = cos A = ߙͳ ܣ ߙͲ ܫ
Since ߣ satisfies the above equation, we have
cosλ = ߙͳ ߙͲ
Putting ߣ = 2 , we get
ߨ ߨ
cos ൌ ߙͳ Ǥ ߙͲ
ʹ ʹ
ߨ
Ͳ ൌ ߙͳ Ǥ ߙͲ
ʹ
Cosߨ = ߙͳ Ǥ ߨ ߙͲ
െͳ ൌ ߙͳ Ǥ ߨ ߙͲ
Form (iii) and (iv) ,we get
ߨ െʹ
ߙͳ . = -1 ͳߙ ൌ Ͳߙ ൌ െͳ ʹ ൌ ͳ
ʹ Ɏ
ߨ ܽߨ
Put ߣ ൌ െ െͳ ൌ ߶ሺߣሻሺͲሻ െ ܾ
ʹ ʹ
ܽߨ
െ ܾ ൌ െͳ ǥ ǥ ǥ ǥ ǥ Ǥ Ǥ ሺܤሻ
ʹ
Solving (A) and (B), we get
ʹ Ͷ
ܽൌ ǡ ܾൌെ
ͷߨ ͷ
͵ߨ ʹ Ͷ
ߣ ൌ ߶ሺߣሻ ൬ߣʹ െ ߣ െ ߨʹ൰ ߣ ൬െ ൰
ʹ ͷߨ ͷ
͵ߨ ʹ Ͷ
ܣ ൌ ߶ሺܣሻ ൬ ʹܣെ ܣെ ߨ ʹ ܫ൰ ܣെ ܫ
ʹ ͷߨ ͷ
ʹ ͷ
ൌ ܣെ ܫ
ͷߨ ͷ
ʹ ߨ ͵ߨ Ͷ ͳ Ͳ െ͵Ȁͷ ͵Ȁͷ
ൌ ቈ ʹ ʹ െ ቂ ቃ ൌ ൨
ͷߨ ߨ ߨ ͷ Ͳ ͳ ʹȀͷ െʹȀͷ
ૠ
16. If ൌ ቂ ቃ find in terms of .
Solution: The characteristic equation is
PROBLEMS
ͺ െͺ െʹ
1. Find the Eigen values and Eigen vectors of the matrix ܣൌ Ͷ െ͵ െʹ൩
͵ െͶ ͳ
ʹ
ǣ͵ǡ ͳ൩ ʹǡͳǡͳ
൦ ൪
ͳ
െ
Ͷ
2. Find the Eigen values and Eigen vectors for ܣൌ ͳ ͵ ʹ൩
െͳ െͷ െʹ
ሾǣɉʹǡ ൌ ሾ͵ǡͳǡ െʹሿԢ ሿ
ʹ ʹ ͳ
3. If ܣൌ ͳ ͵ ͳ൩, find the characteristic vectors of ͵ܣ ܫ
ͳ ʹ ʹ
࢞ ࢞ ࢞ ࢞ ࢞ ࢞ ࢞ ࢞ ࢞
[Ans: By Crammer’s rule ൌ െ ൌ െ ฺ ൌ ൌ ฺ ൌ ൌ
ቚ ቚ ቚ ቚ ቚ ቚ
െ െ
Hence, X= [1, 1, 1]’ is corresponding Eigen vector]
ʹ Ͷ
4. If ܣൌ ቂ ቃ, then find the Eigen values of ܣെͳ ͵ܣ ʹܫ ሾǣ ͳ͵ǡ͵ͳሿ
Ͳ ͵
5. Verify Caley- Hamilton theorem for the matrix A and hence, find ܣെͳ ܣͶ where
ͳ ʹ െʹ
ܣൌ െͳ ͵ Ͳ൩
Ͳ െʹ ͳ
െͷͷ ͳͲͶ ʹͶ
ǣ െʹͲ െͳͷ ͵ʹ൩ ሺܣെͳ ൌ ͳሻ൩
͵ʹ െͶʹ ͳ͵
ʹ െͳ ͳ
6. Find the characteristic equation of the matrix െͳ ʹ െͳ൩ and verify that it is satisfied
ͳ െͳ ʹ
by A and hence, obtain ܣെͳ .
ͳ ͵ ͳ െͳ
ǣ െͳ ൌ ͳ ͵ ͳ ൩
Ͷ
െͳ ͳ ͵
െʹ ͷ Ͷ
7. Find the Eigen values and the corresponding Eigen vectors of the matrix ͷ ͷ ൩
Ͷ ͷ െʹ
show that the Eigen vectors are orthogonal.
ሾǣ െ ͵ǡ െǡͳʹǢ ሾͳǡ െͳǡͳሿԢ ǣ ሾെͳǡͲǡͳሿԢ ǣ ሾͳǡʹǡͳሿԢ ሿ
͵ ͳ
8. Use Cayley-Hemilton theorem to find ʹܣͷ െ ͵ܣͶ ʹܣെ Ͷ ܫwhere ܣቚ ቚ
െͳʹ
ͳͳ ͳ͵ͺ
ቂǣ ቚ ቚቃ.
െͳ͵ͺ ͳʹ
MATRIX THEORY | 57
ͳ ʹ
9. Verify Cayley – Hemilton theorem for ܣൌ ቂ ቃ and evaluate ʹ ʹܣെ ͷ ͵ܣെ ܣ ܫ
ʹ ʹ
ሾǣ Ǥ ͳ ܣ ʹͲ ܫሿ
ͺ െͺ െʹ
10. Show that the matrix ܣൌ Ͷ െ͵ െʹ൩ is diagnosable. Find the transforming matrix and
͵ െͶ ͳ
the diagonal matrix.
ሾܵܰܣǣ Ǥ ͵ǡ ሾʹǡͳǡͳሿ ǡ
ͺ െͺ ʹ
ܯെͳ ܯܣൌ ܣܦͶ െ͵ െʹ൩
͵ െͶ ͳ
ͳ Ͳ Ͳ Ͷ ͵ ʹ
ܦൌ Ͳ ʹ Ͳ൩ ܯൌ ͵ ʹ ͳ൩
Ͳ Ͳ ͵ ʹ ͳ ͳ
ͳ െ െͶ
11. Show that the matrix ܣൌ Ͳ Ͷ ʹ ൩is similar to a diagonal matrix .Also find the
Ͳ െ െ͵
transforming matrix and the diagonal matrix.
ͳ െ െͶ
ǣ Ǥ ǡ ܯെͳ ܯܣൌ ܦǡ ܣൌ Ͳ Ͷ ʹ ൩
Ͳ െ െ͵
Ͳ Ͳ Ͳ ʹ ͳ ʹ
ܦൌ Ͳ ͳ Ͳ൩ ܯൌ െͳ െʹ െʹ൩
Ͳ Ͳ ͳ ʹ ͵ ͵
͵ െͳ ͳ
12. Reduce the following matrix to diagonal form െͳ ͷ െͳ൩
ͳ െͳ ͵
ʹ Ͳ Ͳ
ǣ ܦൌ Ͳ ͵ Ͳ൩
Ͳ Ͳ
Ͷ െͳ
13. Show that A = Ͷ െͳ൩ is derogatory.
െͶ െͶ Ͷ
Ͳ Ͳ Ͳ
ؔ Ͳ Ͳ Ͳ൩ ݂ ሺݔሻ ൌ ʹ ݔെ ͳͷ ͵Ǥ ൩
Ͳ Ͳ Ͳ
ͳ ʹ ͵
14. Show that A = ʹ ͵ Ͷ൩ is non-derogatory.
͵ Ͷ ͷ
ͳ Ͳ Ͳ
15. Shoe that A = ͳ െͳ Ͳ ൩ is derogatory and find its minimal polynomial.
ͳ Ͳ െͳ
ሾǣ ሺሻ ൌ ʹ െ ͳሿ
58 | APPLIED MATHEMATICS - III
ʹ ʹ ͳ
16. Determine the value Eigen value of the associated Eigen vectors for the matrix ͳ ͵ ͳ൩
ͳ ʹ െͳ
ሾǣͳ ʹ െ ͵͵ ൌ Ͳ െ Ͷʹ Ͷ͵ ൌ Ͳܺ͵ ൌ ͳǡ ʹܺݐ݁݃݁ݓ ൌ ͳǤ
ǡ ͷ ܺ ൌ ሾͳǡͳǡͳሿ൧
ͺ െ ʹ
17. Determine Eigen value and Eigen vector for the matrix. െ െͶ൩.
ʹ െͶ ͵
ሾǣ Putting ͵ ൌ ͳwe get ʹ ൌ െʹand thenͳ ൌ ʹǤ Hence, corresponding to the eigen
valueɉ ൌ ͳͷǡwe get The eigen vectorǡ ൌ ሾʹǡ െʹǡͳሿ൧
ͺ െͺ െʹ
18. Find the Eigen values and Eigen vectors of the matrix A = Ͷ െ͵ െʹ൩
͵ െͶ ͳ
[ǣߣ ൌ ͳǡ ʹǡ ͵Ǥሺ݅ሻߣ = [4, 3, 2], (݅݅ሻߣ = [3, 2, 1], ሺ݅݅݅ሻߣ = [2, 1, 1]]
19. Find the Eigen values and Eigen vectors of the matrix. Also verify that the Eigen vectors
ʹ െͳ ͳ
are linearly independent. A = ͳ ʹ െͳ൩
ͳ െͳ ʹ
[ ߣ ൌ ͳǡ ʹǡ ͵Ǥ ሺ݅ሻሾͲǡ ͳǡ ͳሿ, ሺ݅݅ሻሾͳǡ ͳǡ ͳሿ, ሺ݅݅݅ሻሾͳǡ Ͳǡ ͳሿ]
ʹ ʹ ͳ
20. Determine the Eigen values and the associated Eigen vectors for the matrix.ͳ ͵ ͳ൩
ͳ ʹ ʹ
[ǣߣ ൌ ͳǡ ͳǡͷǤ ሺ݅ሻ ͳݔൌ ሾʹǡ െͳǡ ͲሿǢ ሾ ʹݔൌ ሾͳǡ Ͳǡ െͳ], ሺ݅݅ሻ ݔൌ ሾͳǡ ͳǡ ͳሿ]
ʹ ͳ ͳ
21. Find the Eigen values and Eigen vectors corresponding to the matrix.ʹ ͵ ʹ൩
͵ ͵ Ͷ
[ǣߣ ൌ ͳǡ ͳǡ Ǥ ሺ݅ሻ ͳݔൌ ሾͳǡ െͳǡ ͲሿǢ ʹݔൌ ሾͳǡ Ͳǡ െͳሿ; ሺ݅݅ሻ ݔൌ ሾͳǡ ʹǡ ͵ሿ]
ͺ െ ʹ
22. Determine Eigen values and Eigen vectors for the matrix A = െ െͶ൩. Are the
ʹ െͶ ͵
vectors linearly independent ?
[ǣߣ ൌ Ͳǡ ͵ǡ ͳͷǤ ሺ݅ሻ ݔൌ ሾͳǡ ʹǡ ʹሿ; ሺ݅݅ሻ ݔൌ ሾʹǡ ͳǡ െͳሿ; (݅݅݅ሻ ݔൌ ሾʹǡ െʹǡ ͳሿ;
ܾሻ݇͵ ൌ Ͳǡ ݇ʹ ൌ Ͳǡ ݇ͳ ൌ ͲǤ]
͵ െͳ ͳ
23. Find the Eigen values and the Eigen vectors for the matrix. െͳ ͷ െͳ൩
ͳ െͳ ͵
[ǣߣ ൌ ʹǡ ͵ǡ ; ሺ݅ሻ ݔൌ ሾͳǡ Ͳǡ െͳሿ; ሺ݅݅ሻ ݔൌ ሾͳǡ ͳǡ ͳሿ; ሺ݅݅݅ሻ ݔൌ ሾͳǡ െʹǡ ͳሿ]
MATRIX THEORY | 59
െʹ ʹ
24. Find the Eigen values and Eigen vectors of the matrix. െʹ ͵ െͳ൩
ʹ െͳ ͵
[ǣߣ ൌ ʹǡ ʹǡ ͺ; ݅ሻ ͳݔൌ ሾͳǡ ʹǡ ͲሿǢ ʹݔൌ ሾͳǡ Ͳǡ െʹሿ; ݅݅ሻ ͵ݔൌ ሾʹǡ െͳǡ ͳሿ]
͵ െͳ ͳ
25. Find the Eigen values and Eigen vectors for the matrix.െͳ ͵ െͳ൩
ͳ െͳ ͵
[ǣߣ ൌ ʹǡ ʹǡ ͷ; ݅ሻ ͳݔൌ ሾͳǡ ͳǡ ͲሿǢ ʹݔൌ ሾͳǡ Ͳǡ െͳሿ; ݅݅ሻ ͵ݔൌ ሾͳǡ െͳǡ ͳሿ]
Ͷ
26. Find the Eigen values and Eigen vectors for A = ͳ ͵ ʹ൩
െͳ െͷ െʹ
[ǣߣ ൌ ͳǡ ʹǡ ʹ; ሺ݅ሻ ݔൌ ሾͶǡ ͳǡ െ͵ሿ; ሺ݅݅ሻ ݔൌ ሾ͵ǡ ͳǡ െʹሿ]
ʹ ͳ Ͳ
27. Find the Eigen values and the Eigen vectors of the matrix A = Ͳ ʹ ͳ൩.
Ͳ Ͳ ʹ
ൣǣ ݔൌ ሾͳǡ Ͳǡ Ͳሿ൧
ʹ ʹ ͳ
28. If ܣൌ ͳ ͵ ͳ൩ǡ find the characteristic root and characteristic vectors of ͵ܣ ܫ
ͳ ʹ ʹ
ʹ Ͷ ͺ
29. If ߣͳ ǡ ߣʹ ǡ ߣ͵ ܽ݊݀ߣͶ are eigen values of A where ܣൌ ͵ െʹ Ͳ ʹ. Find the eigen
ͳ ͵ ʹ
͵ ʹ ͳ Ͷ
values of adj.A
ሾǣߣͳ ߣʹ ߣ͵ ǡ ߣͳ ߣʹ ߣͶ ǡߣͳ ߣ͵ ߣͶ ǡߣʹ ߣ͵ ߣͶ ሿ
െͳ ʹ ͵
30. If ܣൌ Ͳ ͵ ͷ ൩ǡ find the eigen values of ͵ܣ ͷ ܣ ͺܫ
Ͳ Ͳ െʹ
ሾǤ ǣ ܣെ ͳǡ͵ǡʹ ͵ܣ ͷ ܣ ͺʹܫǡͷͲǡ െͳͲǤ ሿ
31. Verify Cayley- Hamilton Theorem for the matrix A and hence find ܣെͳ ܣͶ
ͳ ʹ െʹ
ܣൌ െͳ ͵ Ͳ ൩ ܣܣെͳ ൌ ͳ
Ͳ െʹ ͳ
ͳ ʹ ͵
32. Find the characteristic equation of the matrix A. where ܣൌ ʹ െͳ Ͷ൩
͵ ͳ െͳ
60 | APPLIED MATHEMATICS - III
QUADRATIC FORM
An homogeneous polynomial of second degree, in ݊ number of unknown or variables is called a
quadratics form ሺܳܨሻ in ݊ variables.
For Example: ܳ ൌ ʹͳݔ ʹ ʹʹݔ ͵ ʹ͵ݔ ʹ ͵ݔ ʹݔെ ʹ ͳݔ ͵ݔ Ͷ ʹݔ ͳݔis a quadratic form in three
variables ͳݔǡ ʹݔand ͵ݔ.
Quadratic form ܳ can be written as a product of three matrices ܺ Ԣ ǡ ܣand ܺ i.e. ܳ ൌ ܺ Ԣ ܺܣ,
where
(a) ܺ is a column matrix of unknowns or variables.
(b) ܺ Ԣ is the transpose of ݔ.
(c) ܣis a symmetric matrix of order equal to the number of variables in quadratic form.
Symmetric matrix ܣis called the matrix of the quadratic form.
Note: (i) The diagonal elements of ܣare the coefficients of square terms in ܳ. (ii) Non diagonal
elements of ܣare half the coefficients of product terms in ܳ.
SOLVED EXAMPLE
1. Write the matrix of quadratic forms of the following:
(i) ܳ ൌ ʹͳݔെ ʹ ʹʹݔ ͵ ʹ͵ݔെ Ͷ ʹݔ ͳݔ ͷ ͵ݔ ʹݔ ͵ݔ ͳݔ
Solution: The matrix ܣof quadratic form is
ͳ െʹ ͵
ܣൌ െʹ െʹ ͷȀʹ൩
͵ ͷȀʹ ͵
(ii) ܳ ൌ ʹ ʹͳݔ Ͷ ʹʹݔ ʹ ͵ݔ ͳݔ ͺ͵ݔ ʹݔ
Solution: This can be written as
ܳ ൌ ʹ ʹͳݔ Ͷ ʹʹݔ Ͳ ʹ͵ݔ Ͳ ʹݔ ͳݔ ʹ ͵ݔ ͳݔ ͺ͵ݔ ʹݔ
Therefore the matrix ܣof quadratic form is
ʹ Ͳ ͳ
ܣൌ ͲͶͶ൩
ͳ Ͷ Ͳ
(iii) ܳ ൌ ͵ ʹͳݔ Ͷ ʹʹݔ ͷʹ͵ݔ
Solution: Therefore the matrix ܣof quadratic form is
͵ Ͳ Ͳ
ܣൌ ͲͶͲ൩
Ͳ Ͳ ͷ
(iv) ܳ ൌ ܽ ʹ ݔ ʹܾ ݕݔ ܾʹ ݕ
Solution: Therefore the matrix ܣof quadratic form is
ܽ ݄
ܣൌ ቂ ቃ
݄ ܾ
MATRIX THEORY | 61
2. Write the quadratic form where the matrix of quadratic form is ൌ ૢ ൩
ૢ െ
Solution: The quadratic form is
ܳ ൌ ʹͳݔ Ͷ ʹʹݔെ ͵ ʹ͵ݔ Ͷ ʹݔ ͳݔ ͳͺ ͵ݔ ʹݔ ͳݔ ͵ݔ
Linear Transformations
Consider two sets of variables ͳݔǡ ʹݔǡ ǥ ǥ Ǥ ǡ ݊ݔand ͳݕǡ ʹݕǡ ǥ ǥ ǡ ݊ݕ.
Let these variables be related by the following linearly equations.
ͳݔൌ ͳݕ ͳͳ ʹݕ ʹͳ ڮǥ Ǥ ݊ݕ ݊ͳ
ʹݔൌ ͳݕ ͳʹ ʹݕ ʹʹ ڮǥ ݊ݕ ݊ʹ
͵ݔൌ ͳݕ ͳ͵ ʹݕ ʹ͵ ڮǥ Ǥ Ǥ ݊ݕ ݊͵
ڭ
ڭ
݊ݔൌ ͳݕ ͳ݊ ʹݕ ʹ݊ ڮǥ ǥ ݊ݕ ݊݊
The above ݊ equations can be expressed as a single matrix equation. ܺ ൌ ܻܲ,
ͳݔ ݊ͳ ڮ ʹͳ ͳͳ ͳݕ
ʹݔ ݊ʹ ڮ ʹʹ ͳʹ ʹݕ
where ܺ ൌ ǡ ܲ ൌ ǡ ܻ ൌ
ڭ ڭ ڭ ڭ ڭ ڭ
݊ݔ ݊݊ ڮ ʹ݊ ͳ݊ ݊ݕ
The transformation from ’ݔs to ’ݕs by the above equations is called a linear transformation. The
linear transformation is singular or non singular according as the matrix ܲ is singular or non
singular.
Linear Transformation of a Quadratic form
Consider a quadratic form ܳ ൌ ܺ Ԣ ܺܣand a non singular linear transformation given by ܺ ൌ ܻܲ
ܳ ൌ ܺ Ԣ ܺܣൌ ሺܻܲሻԢ ܣሺܻܲሻ
ൌ ܻ Ԣ ሺܲԢ ܲܣሻܻ ൌ ܻ Ԣ ܻܦ
where ܦൌ ܲԢ ܲܣ
where ܣis a matrix of rank ݎthen there exists a non singular linear transformation ܺ ൌ ܻܲ
which transforms the given quadratic form to sum of ݎsquare terms, i.e.
ܾͳ ʹͳݕ ܾʹ ʹʹݕ ڮǥ ܾ݊ ʹ݊ݕ
The Quadratic form ܻ Ԣ ܻܦis called a linear transform of the quadratic form ܺ Ԣ ܺܣ. This new
form ܳ ൌ ܻ Ԣ ( ܻܦi.e. the linear transform of the quadratic form) has been expressed as the sum
and difference of the squares of new variables ’ݕs. This form is called canonical form.
[Thus if by any non singular linear transformation a real quadratic form can be expressed as a
sum and difference of the squares of the new variables then this later expression is called the
canonical form of the given form]
62 | APPLIED MATHEMATICS - III
(a) Rank of a quadratic form: Rank of the diagonal matrix ܦis called the rank of the
quadratic form. It is also equal to the number of non-zero terms in its canonical form. It is
denoted by ݎ.
(b) Index of a quadratic form ሺሻ: The number of positive squares in the canonical form is
called the index of the quadratic form. It is denoted by .
(c) Signature of a quadratic form ሺ࢙ሻ: It is the excess number of positive squares over
negative squares in canonical form. Since there are in all ݎterms and of them are
positive then there will be ݎെ negative terms and hence the difference between the
positive and negative terms is called signature of the quadratic form. Thus
ݏൌ െ ሺ ݎെ ሻ ൌ ʹ െ ݎ
Classification of quadratic forms
Value class of a quadratic form ܳ ൌ ܺ Ԣ ܺܣൌ ܻ Ԣ ܻܦ
(a) Positive definite: If ݎൌ ݊ and ൌ ݊
i.e. the number of positive squares is equal to the number of variables and also equal to
the rank or in other words if all the diagonal elements of matrix ܦare ( ݁ݒnot zero) then
the quadratic form is called positive definite. i.e. quadratic form can be reduced to form
ܳ ൌ ܾͳ ʹͳݕ ܾʹ ʹʹݕ ڮǥ ǥ ܾ݊ ʹ݊ݕሺܾ݊ Ͳሻ
(b) Negative definite: If ݎൌ ݊ and ൌ Ͳ
i.e. the number of negative squares is equal to the number of variables and also equal to
the rank or in other words if all the diagonal elements of matrix ܦare െ ݁ݒthen the
quadratic form is called negative definite. i.e. quadratic form can be reduced to form
ܳ ൌ െܾͳ ʹͳݕെ ܾʹ ʹʹݕെ ڮǥ ǥ െ ܾ݊ ʹ݊ݕ
(c) Positive semi definite: If ݎ൏ ݊ and ൌ ݎ
i.e. if the number of positive squares is equal to the rank of the matrix of the quadratic
form which is less than the number of variables ݊ or in other words if ݎdiagonal elements
of the matrix ܦare positive and remaining ሺ݊ െ ݎሻ elements are zero i.e. then the
quadratic form is called positive semi definite. i.e. quadratic form can be reduced to form
ܳ ൌ ܾͳ ʹͳݕ ܾʹ ʹʹݕ ڮǥ ǥ ܾ ʹݎݕ ݎሺܾ ݎ Ͳሻ
(d) Negative semi definite: If ݎ൏ ݊ and ൌ Ͳ
i.e. if the number of negative squares is equal to the rank of the matrix of the quadratic
form which is less than the number of variables ݊ or in other words if ݎdiagonal elements
of the matrix ܦare negative and remaining ሺ݊ െ ݎሻ elements are zero i.e. then the
quadratic form is called negative semi definite. i.e. quadratic form can be reduced to form
ܳ ൌ െܾͳ ʹͳݕെ ܾʹ ʹʹݕെ ڮǥ ǥ െ ܾ ʹݎݕ ݎሺܾ ݎ Ͳሻ
(e) Indefinite: If the matrix ܦhas some diagonal elements ݁ݒand some diagonal elements
Ȃ ݁ݒ, or zero, the quadratic form is called indefinite.
The above five classes of real quadratic forms are called value classes of real quadratic forms.
MATRIX THEORY | 63
SOLVED EXAMPLE
1. Reduce the quadratic form ࢞ ࢞ െ ࢞ െ ૡ࢞ ࢞ െ ࢞ ࢞ ࢞ ࢞ to normal form
through congruent transformations. Also find its rank, signature and value class.
Solution: The quadratic form can be written as
ʹ െʹ ܺͳ ʹ െʹ
ሾ ͵ݔ ʹݔ ͳݔሿ ͳ െͶ൩ ܺʹ ൩ ܣ ൌ ͳ െͶ൩
െʹ െͶ െ͵ ܺ͵ െʹ െͶ െ͵
We write A = I A I
ʹ െʹ ͳ Ͳ Ͳ ͳ Ͳ Ͳ
ͳ െͶ൩ ൌ Ͳ ͳ Ͳ൩ AͲ ͳ Ͳ൩
െʹ െͶ െ͵ Ͳ Ͳ ͳ Ͳ Ͳ ͳ
ܴʹ െ ͵ܴͳ
ܥെ ͵ʹ ͳܥ Ͳ Ͳ ͳ Ͳ Ͳ ͳ െ͵ ͳ
By ʹ ʹ ൩ ൌ െ͵ ͳ Ͳ ൩ ܣͲ Ͳ൩
ܴ͵ ܴͳ Ͳ െͳ ͳ
Ͳ ʹ െͷ ͳ Ͳ ͳ Ͳ Ͳ ͳ
͵ܥ ͳܥ
ͳͳ
ʹ ʹ Ͳ Ͳ ͳ Ͳ Ͳ ͳ െ͵
ܴ͵ ܴʹ ͳ
By ͳ
ʹ Ͳ െͳ Ͳ ൌ െ͵
ͺͳ ͳͳ
ͳ
ʹ
Ͳ A ൦
Ͳ ͳ
ʹ൪
͵ܥ Ͳ ʹܥ ʹ ͳ
ͳ ͳ ͳ ͳ Ͳ Ͳ ͳ
ͳ ͳ ͳ ξͳ ξͳ
By ܴͳ ǡ ܥǢ ܥǢ ܴ͵ ǡ ͵ܥ
ξʹ ξʹ ͳ ξͳ ʹ ͻ ͻ
ͳ ͳ ͵ ͳͳ
ۍ Ͳ Ͳ ۍ ې െ ې
ێξʹ ێ ۑξʹ ξͳ ͻξͳۑ
ͳ Ͳ Ͳ ێ ͵ ͳ ۑ ͳ ʹ ۑ
Ͳ െͳ Ͳ൩ ൌ ێെ Ͳ Ͳ ێێ ܣ ۑ
Ͳ Ͳ ͳ
ξͳ ξͳ ξͳ ͻξͳۑ
ێ ێ ۑ ۑ
ͳͳ ێ ʹ ξͳێ ۑ ξͳ ۑ
Ͳ Ͳ
ͻ ۏξͳ ͻξͳ ͻ ۏ ے ͻ ے
Thus, B = ܲԢ ܲܣwhere B is diagonal matrix.
This means the quadratic form ܺ Ԣ ܺܣൌ ʹ ʹͳݔ ʹʹݔെ ͵ ʹ͵ݔെ ͺ ͵ݔ ʹݔെ Ͷ ͳݔ ͵ݔ ͳʹ ʹݔ ͳݔwill be
transformed to ܻ Ԣ ܻܤൌ ʹͳݕെ ʹʹݕ ʹ͵ݕby transformation X = PY.
ͳ ͳ ͳͳ ͳ ʹ ξͳ
ฺ ͳݔൌ ͳݕെ ʹݕ ͵ݕǢ ʹݔൌ ʹݕ ͵ݕǢ ͵ݔൌ ݕ
ξʹ ξͳ ͻξ ξͳ ͻξͳ ͻ ͵
The rank = 3
Signature = Difference between positive squares and negative squares.
=2–1=1
Since some diagonal elements are positive ,some are negative ,the value class is indefinite.
64 | APPLIED MATHEMATICS - III
Now rank ݎof the given quadratic form ൌ no. of non zero terms in its canonical form ൌ ͵ Index
ൌ no. of positive squares in canonical form ൌ ʹ.
Signature of the given quadratic form ൌ excess number of positive squares over number of
negative squares in its canonical form ൌ ʹ െ ݎൌ Ͷ െ ͵ ൌ ͳ.
For value class,
ݎൌ݊ൌ͵
ൌʹ
The value class of the given quadratic form is indefinite.
Now the linear transformation ܺ ൌ ܻܲ
ͳݔ ͳ െͳ ͵ ͳݕ
ʹݔ൩ ൌ Ͳ ͳ െʹ൩ ʹݕ൩
͵ݔ Ͳ Ͳ ͳ ͵ݕ
ͳݔൌ ͳݕെ ʹݕ ͵͵ݕ
ʹݔൌ Ͳ ͳݕ ʹݕെ ʹ͵ݕ
͵ݔൌ Ͳ ͳݕ Ͳ ʹݕ ͵ݕ
ܳ ൌ ʹͳݕ ʹʹݕെ ʹʹ͵ݕ
(i) For ࡽ positive: Take ͳݕൌ ʹǡ ʹݕൌ ͳǡ ͵ݕൌ ͳ
ܳ ൌ Ͷ ͳ െ ʹ ൌ ͵
ͳݔ ൌ ʹ െ ͳ ͵ ൌ Ͷ
ฺ ʹݔൌ ͳ െ ʹ ൌ െͳ
ฺ ͵ݔൌ ͳ ൌ ͳ
Therefore taking ͳݔൌ Ͷǡ ʹݔൌ െͳǡ ͵ݔൌ ͳ
ܳ ൌ ሺͶሻʹ ʹሺെͳሻʹ ͵ሺͳሻʹ ʹሺെͳሻሺͳሻ െ ʹሺͳሻሺͶሻ ʹሺͶሻሺെͳሻ
ൌ ͳ ʹ ͵ െ ʹ െ ͺ െ ͺ ൌ ͵ which is a positive value.
(ii) For ࡽ negative: Take ͳݕൌ ͳǡ ʹݕൌ ͳǡ ͵ݕൌ ʹ
ܳ ൌ ͳ ͳ െ ͺ ൌ െ
ฺ ͳݔൌ ͳ െ ͳ ൌ
ฺ ʹݔൌ ͳ െ Ͷ ൌ െ͵
ฺ ͵ݔൌ ʹ
ܳൌ ሺሻʹ ʹሺെ͵ሻʹ ͵ሺʹሻʹ ʹሺെ͵ሻሺʹሻ െ ʹሺʹሻሺሻ ʹሺሻሺെ͵ሻ
ൌ ͵ ͳͺ ͳʹ െ ͳʹ െ ʹͶ െ ͵
ൌ െ
66 | APPLIED MATHEMATICS - III
͵ Ͳ Ͳ ͳ Ͳ Ͳ ͳ െ
ͳ
െ
ʹ͵
ܴ͵ ܴʹ Ͳ
Ͳ൪ ൌ ൦ െ
ͳ
ͳ Ͳ൪ ܣ൦ ͵ ͳͶ
By ͵ ൦ ͵ ͵
Ͳ ͳ
͵ ൪
͵ܥ Ͳ ʹܥ Ͳ
ͳ ʹ͵ ͵
െ ͳ Ͳ Ͳ ͳ
ͳͶ ͳͶ
ͳ ʹ͵ ͵
ฺ ͳݔൌ ͳݕെ ʹݕെ ͵ݕǡ ʹݔൌ ʹݕ ͵ݕǡ ͵ݔൌ ͵ݕǤ
͵ ͳͶ
MATRIX THEORY | 67
4. Show that the quadratic form ࢞ ૠ࢟ ࢠ െ ࢞࢟ െ ࢟ࢠ ૡࢠ࢞ in three
variables is positive semi definite. Find a non zero set of values of ࢞ǡ ࢟ǡ ࢠ for which the
form assumes positive value.
െͳͲ Ͷ ݔ
Solution: We have quadratic form ܳ ൌ ሾݖ ݕ ݔሿ െͳͲ ͳ െ൩ ቈݕ
Ͷ െ ͵ ݖ
െͳͲ Ͷ
The matrix ܣof the quadratic form is ܣൌ െͳͲ ͳ െ൩
Ͷ െ ͵
Let ܣൌ ͵ܫܣ ͵ܫ
െͳͲ Ͷ ͳ Ͳ Ͳ ͳ Ͳ Ͳ
ฺ െͳͲ ͳ െ൩ ൌ ͲͳͲ൩ ܣͲͳͲ൩
Ͷ െ ͵ Ͳ Ͳ ͳ Ͳ Ͳ ͳ
െ͵Ͳ Ͷ ͳ Ͳ Ͳ ͳ Ͳ Ͳ
By ͵ܴʹ and ͵ ʹܥെ͵Ͳ ͳͷ͵ െʹͳ൩ ൌ Ͳ͵Ͳ൩ ܣͲ͵Ͳ൩
Ͷ െʹͳ ͵ Ͳ Ͳ ͳ Ͳ Ͳ ͳ
െ͵Ͳ ͳʹ ͳ Ͳ Ͳ ͳ Ͳ Ͳ
By ͵ܴ͵ and ͵ ͵ܥെ͵Ͳ ͳͷ͵ െ͵൩ ൌ Ͳ͵Ͳ൩ ܣͲ͵Ͳ൩
ͳʹ െ͵ ʹ Ͳ Ͳ ͵ Ͳ Ͳ ͵
െ͵Ͳ ͳʹ ͳ Ͳ Ͳ ͳ Ͳ Ͳ
By ܴʹ ͷܴͳ Ͳ ͵ െ͵൩ ൌ ͷ͵Ͳ൩ ܣͲ͵Ͳ൩
ͳʹ െ͵ ʹ Ͳ Ͳ ͵ Ͳ Ͳ ͵
Ͳ ͳʹ ͳ Ͳ Ͳ ͳ ͷ Ͳ
By ʹܥ ͷ ͳܥ Ͳ ͵ െ͵൩ ൌ ͷ͵Ͳ൩ ܣͲ͵Ͳ൩
ͳʹ െ͵ ʹ Ͳ Ͳ ͵ Ͳ Ͳ ͵
Ͳ ͳʹ ͳ Ͳ Ͳ ͳ ͷ Ͳ
By ܴ͵ െ ʹܴͳ Ͳ ͵ െ͵൩ ൌ ͷ ͵Ͳ൩ ܣͲ͵Ͳ൩
Ͳ െ͵ ͵ െʹ Ͳ ͵ Ͳ Ͳ ͵
Ͳ Ͳ ͳ Ͳ Ͳ ͳ ͷ െʹ
By ͵ܥെ ʹ ͳܥͲ ͵ െ͵൩ ൌ ͷ ͵Ͳ൩ ܣͲ͵ Ͳ ൩
Ͳ െ͵ ͵ െʹ Ͳ ͵ Ͳ Ͳ ͵
Ͳ Ͳ ͳ Ͳ Ͳ ͳ ͷ െʹ
By ܴ͵ ܴʹ Ͳ͵െ͵൩ ൌ ͷ͵Ͳ൩ ܣͲ͵ Ͳ ൩
Ͳ Ͳ Ͳ ͵ ͵ ͵ Ͳ Ͳ ͵
Ͳ Ͳ ͳ Ͳ Ͳ ͳ ͷ ͵
By ͵ܥ ʹܥͲ͵Ͳ൩ ൌ ͷ͵Ͳ൩ ܣͲ͵͵൩
Ͳ Ͳ Ͳ ͵ ͵ ͵ Ͳ Ͳ ͵
ൌ ܲԢ ܲܣ
ͳ ͷ ͵
ܲ ൌ Ͳ͵͵൩
Ͳ Ͳ ͵
68 | APPLIED MATHEMATICS - III
Ͳ Ͳ ͳݕ
ܳ ൌ ܻ ܻܦൌ ሾͳݕ
Ԣ ʹݕ ͵ݕሿ Ͳ͵Ͳ൩ ʹݕ൩
Ͳ Ͳ Ͳ ͵ݕ
ܳ ൌ ʹͳݕ ͵ ʹʹݕ Ͳʹ͵ݕ
This is now expressed as sum and difference of the squares of new variable ͳݕǡ ʹݕǡ ͵ݕ. This is
called canonical form.
Now rank ݎof the given quadratic form ൌ no. of non zero terms in its canonical form ൌ ʹ index
ൌ no. of positive squares in normal form ൌ ʹ. Signature of the given quadratic form is the
excess number of positive squares over number of negative squares in its canonical form
ൌ ʹ െ ݎൌ Ͷ െ ʹ ൌ ʹ
The value class of the given quadratic form is positive semi definite.
Now ܳ ൌ ʹͳݕ ͵ ʹʹݕ Ͳʹ͵ݕ
Now the linear transformation as ܺ ൌ ܻܲ
ͳ ݔͷ ͵ ͳݕ
ݕ ݕ
ቈ Ͳ͵͵൩ ൌ ʹ ൩
͵ Ͳ Ͳ ݖ ͵ݕ
ݔ ൌ ͳݕ ͷ ʹݕ ͵͵ݕ
ݕൌ Ͳ ͳݕ ͵ ʹݕ ͵͵ݕ
ݖൌ Ͳ ͳݕ Ͳ ʹݕ ͵͵ݕ
For ܳ to be positive
ͳݕൌ ͳǡ ʹݕൌ ͳǡ ͵ݕൌ Ͳ
ܳ ൌ ሺͳሻʹ ͵ሺͳሻʹ െ Ͳ ൌ ͻ
ݔ ൌ ǡ ݕൌ ͵ǡ ݖൌ Ͳ
Putting these values in the given quadratic form,
ܳ ൌ ሺሻʹ ͳሺ͵ሻʹ െ ʹͶሺሻሺ͵ሻ
ൌ ͻǡሾሿ
5. Reduce the following quadratic form to canonical form by
(a) congruent transformation
(b) orthogonal transformation
State the transformation matrix in each case. Also find the rank, index and signature
of quadratic form and class value.
ࡽ ൌ ࢞ ࢞ ࢞ െ ࢞ ࢞ ࢞ ࢞ െ ࢞ ࢞
Solution: The quadratic form can be expressed as
െʹ ʹ ͳݔ
ܳ ൌ ܺ Ԣ ܺܣൌ ሾͳݔ ʹݔ ͵ݔሿ െʹ ͵ െͳ൩ ʹݔ൩
ʹ െͳ ͵ ͵ݔ
MATRIX THEORY | 69
ߣ͵ െ ͳʹߣʹ ͵ߣ െ ͵ʹ ൌ Ͳ
ሺߣ െ ʹሻሺߣ െ ʹሻሺߣ െ ͺሻ ൌ Ͳ
ߣͳ ൌ ʹǡ ߣʹ ൌ ʹǡ ߣ͵ ൌ ͺ
െʹ െʹ ʹ ͳݔ Ͳ
For ߣͳ ൌ ͺ, the eigen vectors given by െʹെͷെͳ൩ ʹݔ൩ ൌ Ͳ൩
ʹ െͳ െͷ ͵ݔ Ͳ
Let us obtain the rank of the matrix
െʹ െʹ ʹ
ܣെ ߣ ܫൌ െʹെͷെͳ൩
ʹ െͳ െͷ
െʹ െʹ ʹ
ܴʹ െ ܴͳ ǡ ܴ͵ ܴͳ ̱ Ͳ െ͵െ͵൩
Ͳ െ͵ െ͵
െʹ െʹ ʹ
ܴ͵ െ ܴʹ ̱ Ͳ െ͵െ͵൩
Ͳ Ͳ Ͳ
ฺ െʹ ͳݔെ ʹ ʹݔ ʹ ͵ݔൌ Ͳ
ฺ Ͳ ͳݔെ ͵ ʹݔെ ͵ ͵ݔൌ Ͳ
ͳݔ െʹݔ ͵ݔ
ൌ ൌ
െʹ ʹ െʹ ʹ െʹ െͳ
ቚ ቚ ቚ ቚ ቚ ቚ
െ͵ െ͵ Ͳ െ͵ Ͳ െ͵
ͳݔ ʹݔ ͵ݔ
ฺ ൌ ൌ
ͳʹ െ
ͳݔ ʹݔ ͵ݔ
ฺ ൌ ൌ
ʹ െͳ ͳ
ʹ
ฺ ܺͳ ൌ െͳ൩
ͳ
For ߣʹ ൌ ʹ,
To find rank of matrix ܣെ ߣܫ
ሺ ܣെ ߣܫሻܺ ൌ Ͳ
Ͷ െʹ ʹ ͳݔ Ͳ
Gives െʹ ͳ െͳ൩ ʹݔ൩ ൌ Ͳ൩
ʹ െͳ ͳ ͵ݔ Ͳ
MATRIX THEORY | 71
Ͷ െʹ ʹ
െʹ ͳ െͳ൩
ʹ െͳ ͳ
Ͷ െʹ ʹ
ܴ͵ ܴʹ ݃݅ ݏ݁ݒെʹ ͳ െͳ൩
Ͳ Ͳ Ͳ
ͳ Ͷ െʹ ʹ
ܴʹ ܴͳ ݃݅ ݏ݁ݒͲ Ͳ Ͳ൩
ʹ
Ͳ Ͳ Ͳ
Rank is ͳ.
G.M. = no. of unknown rank ൌ ͵ െ ͳ ൌ ʹ
Ͷ െʹ ʹ ͳݔ Ͳ
Ͳ Ͳ Ͳ൩ ʹݔ൩ ൌ Ͳ൩
Ͳ Ͳ Ͳ ͵ݔ Ͳ
Ͷ ͳݔെ ʹ ʹݔ ʹ ͵ݔൌ Ͳ
ฺ ʹ ͳݔെ ʹݔ ͵ݔൌ Ͳ
Putting ʹݔൌ Ͳǡ ͳݔൌ ͳǡ ͵ݔൌ െʹ
ͳ
ܺʹ ൌ Ͳ ൩
െʹ
݈
Now eigen vectors of ܺ͵ ൌ ݉൩ so chosen that ܺʹ ܺ͵ ǡ ܺͳ ܺ͵ .
݊
݈ Ͳ݉ െ ʹ݊ ൌ Ͳ
ฺ ʹ݈ െ ݉ ݊ ൌ Ͳ
By Cramer’s rule,
݈ െ݉ ݊
ൌ ൌ
Ͳ െʹ ͳ െʹ ͳ Ͳ
ቚ ቚ ቚ ቚ ቚ ቚ
െͳ ͳ ʹ ͳ ʹ െͳ
݈ െ݉ ݊
ฺ ൌ ൌ
െʹ െͷ െͳ
െʹ ʹ
ܺ͵ ൌ െͷ൩ ൌ ͷ൩
െͳ ͳ
The matrix of eigen vectors is
ʹ ͳ ʹ
ܯൌ െͳ Ͳ ͷ൩
ͳ െʹ ͳ
Normalizing the column of ܯwe get the orthogonal transformation matrix as
72 | APPLIED MATHEMATICS - IV
ʹȀξͳȀξͷ ʹȀξ͵Ͳ
ܲ ൌ ൦െͳȀξ Ͳ ͷȀξ͵Ͳ൪ ܰ݁ݐሺܲܲԢ ൌ ܫሻ ܲ െͳ ൌ ܲԢ
ͳȀξ െʹȀξͷ ͳȀξ͵Ͳ
The orthogonal transformation ܺ ൌ ܻܲ is
ͳݔ ʹȀξ ͳȀξͷ ʹȀξ͵Ͳ ͳݕ
ʹݔ൩ ൌ ൦െͳȀξ Ͳ ͷȀξ͵Ͳ൪ ʹݕ൩
͵ݔ ͵ݕ
ͳȀξ െʹȀξͷ ͳȀξ͵Ͳ
ʹ ͳ ʹ
ͳݔൌ ͳݕ ʹݕ
͵ݕ
ξ ξͷ ξ͵Ͳ
ͳ ͷ
ʹݔൌ െ ͳݕ Ͳ ʹݕ ͵ݕ
ξͲ
ͳ ʹ ͳ
͵ݔൌ ͳݕെ ʹݕ
͵ݕ
ξ ξͷ ξ͵Ͳ
The canonical form of quadratic form under orthogonal transformation is
ͺ Ͳ Ͳ ͳݕ
ܳ ൌ ܻ Ԣ ܻܦൌ ሾ ͵ݕ ʹݕ ͳݕሿ ͲʹͲ൩ ʹݕ൩ ൌ ͺ ʹͳݕ ʹ ʹʹݕ ʹʹ͵ݕ
Ͳ Ͳ ʹ ͵ݕ
PROBLEMS
1. Reduce the following quadratic forms to diagonal form through congruent transformations.
(a) ʹ ݔ ʹ ʹ ݕ ʹ ݖെ Ͷ ݕݔ ͺݖݔ [Ans: ʹݑെ ʹ ʹ ݒ ͻ ʹ ݓሿ
ͳ
(b) ʹ ݔ ʹ ʹ ݕ ʹ ʹ ݖെ ʹ ݕݔെ ʹ ݖݕ ݔݖ [Ans: ʹݑ ʹ ݒ ʹ ݓሿ
ʹ
ʹ ʹ ʹ ʹ ͳͳ ʹ
(c) ݔ ʹ ݕെ ͵ ݖ ͷ ݓെ Ͷ ݕݔ ͺ ݖݕ ʹ ݓݕെ ʹݔݖ [Ans: െ ʹ ݍെ ʹ ʹ ݎ
ʹ ʹ
ݏሿ
ʹ
2. Reduce the quadratic form ͷ ʹͳݔ ʹ ʹʹݔ ͳͲ ʹ͵ݔ ʹݔ ͳݔ Ͷ ͵ݔ ʹݔ ͳͶ ͳݔ ͵ݔto normal
form Show that positive semi-definite and find non zero set of values of ܺͳ ǡ ܺʹ ǡ ܺ͵ which
will make the quadratic form zero.
3. Reduce the following quadratic forms to canonical form by congruent transformations.
State also form of the quadratic form.
(a) ʹ ݔെ ʹ ʹ ݕ ͵ ʹ ݖെ Ͷ ݖݕ ݔݖ
(b) ʹͳݔ ʹ ʹʹݔ ͵ ʹ͵ݔ ʹ ʹݔ ͳݔെ ʹ ͵ݔ ͳݔ ʹ͵ݔ ʹݔ
ͳ ʹ ͵
4. Find the orthogonal matrix which will diagonalise the matrix ܣൌ ʹ Ͷ ൩
͵ ͻ
Ԣ ʹ ʹ
5. Reduce the quadratic form ܺ ܺܣൌ ͵ ͳݔ ͵ ͵ݔ Ͷ ʹݔ ͳݔ ͺ ͵ݔ ͳݔ Ͷ ͵ݔ ʹݔto the form
ߣͳ ʹ ͳݕ ߣʹ ʹ ʹݕ ߣͳ ͳߣ݁ݎ݄݁ݓ ʹ ͵ݕǡ ߣʹ ǡ ߣ͵ are eigen values of A. Hence, deduce that ܺ Ԣ ܺܣ
is indefinite and find ܺ ് Ͳ such that ܺ Ԣ =ܺܣ0
6. Reduce the quadratic form ʹ ݕݔ ʹ ݖݕ ʹ ݔݖby orthogonal reduction to diagonal form n