5.1.

1 A function f (x) is expanded in a series of orthonormal functions

∞
f ( x ) = ∑ an φ n ( x ) .
n =0

Show that the series expansion is unique for a given set ofφ n ( x ) . The functions φ n ( x )
are being taken here as the basis vectors in an infinite-dimensional Hilbert space.
Answer:
b

Using orthogonality the ¿ ∅ n∨f >¿ a n=∫ ω ( x ) f ( x ) ∅ n ( x ) dx are derived from f and
a

therefore unique.

5.1.2 A function f ( x) is represented by a finite set of basis functions φ n ( x )

n
f ( x )=∑ c i φ i ( x ) .
i=1

Show that the components ci are unique, that no different set exists c i. Note. Your
basis functions are automatically linearly independent. They are not necessarily
orthogonal.
Answer:

If f ( x )=∑ Ci ϕ i ( x )=¿ ∑ c j ϕ j ( x ) ¿ then ∑ ( c i−c 'i ) ϕi=0 Let c k−c 'k ≠ 0 be the first non-
'

i j i

1
; ∑
'
zero term. Then ϕ k = (c i−c k )ϕi .
ck −c k i>k

n −1
A function f (x) is approximated by a power series ∑ ci x over the interval [0, 1].
i
5.1.3
i=0

Show that minimizing the mean square error leads to a set of linear equations
Ac=b
Where
1
1
Aii =∫ x
i+ j
dx= ,i , j=0,1,2, … , n−1
0 i+ j+1
And
1
b i=∫ x fdxi=0,1,2, … , n−1
i

Note. The Aij are the elements of the Hilbert matrix of order n. The determinant of this
Hilbert matrix is a rapidly decreasing function of n. For n = 5, detA = 3.7 × 10-12 and
the set of equations Ac = b is becoming ill-conditioned and unstable.
Answer:
 n−1
For f ( x )=∑ c i x we have
i

i=0

1 1 n−1
ci
b j=∫ x f ( x ) dx=∑ c i∫ x dx=∑
j i+j
= Aij ci
0 i 0 1=0 i+ j+1

This results also from minimizing the mean square error

∫¿¿¿
0

5.1.4 In place of the expansion of a function f (x) given by

∞
f ( x )=∑ c n φn ( x ) .
n =0

With
b
a n=∫ F ( x ) φn ( x ) w ( x ) dx .
a

take the finite series approximation

∞
f ( x ) ≈ ∑ c n φn ( x ) .
n=0

Show that the mean square error

[ ]
b m 2

∫ f ( x )−∑ c n φn ( x ) . w ( x ) dx
a n=0

is minimized by takingc n=an. Note. The values of the coefficients are independent of
the number of terms in the finite series. This independence is a consequence of
orthogonality and would not hold for a least-squares fit using powers of x.
Answer:
From
b
∂
0= ∫¿ ¿
∂ cl a

[ ]
a m
=2∫ f ( x )−∑ c n ϕn (x ) ϕ n w ( x ) dx
b n=0

We obtain c n∫ f ( x ) ∅n ( x ) w ( x ) dx .
b

5.1.5 From Example 5.1.6,

 { }
h
0 < x< π ∞
2h sin ( 2 n+1 ) x
f (x) 2 = ∑
−h π n=0 2n+ 1
−π < x <0
2
(a)show that
π ∞

∫ [f ( x ) ] dx= π2 h2 ∑ (2 n+1)−2
2

−π n=0

For a finite upper limit this would be Bessel’s inequality. For the upper limit ∞, this is
Parseval’s identity
(b) Verify
∞
π 2 4 h2
2
h= ∑ (2 n+ 1)−2
π n=0
by evaluating the series. Hint. The series can be expressed in terms of the Riemann
zeta function ζ (2) π 2 /6
Answer:
(a) and (b)
π 2 ∞

∫ f (x)2 dx ¿ ( h2 ) 2 π ∑ (2 m+ 1)(2
1
n+1)
−π m , n=0

π
× ∫ sin ( 2m+1 ) x sin ( 2 n+1 ) x dx
−π

∞ 2
1 3 π
Using ∑ = ζ ( 2 ) ,we reach ζ ( 2 ) =
n=0 ( 2n+ 1)
2
4 6

5.1.6 Derive the Schwarz inequality from the identity

[ ]
b 2 b b

∫ f ( x ) g(x ) dx =∫ [ f ( x ) ] dx ∫ [g ( x ) ]2 dx
2

a a a

b b
−1
∫
2 a
dx ∫ dy [f ( x ) g ( y )−f ( y ) g ( x )] 2
a

Jawab :
b b
1
¿ ⟨ f ∨g ⟩∨¿ =⟨ f ⟩⟨ g ⟩ − ∫ ∫ ¿ f ( x ) g ( y )−f ( y ) g ( x ) ∨¿ dxdy ¿ ¿ implies
2 2 2 2

2a a
1
¿ ⟨ f ∨g ⟩∨¿2 ≤ ⟨ f 2 ⟩ ⟨ g2 ⟩− ¿ because the double integral is nonnegative.
2

5.1.7 ⟨
Starting from I= f −∑ ai φi f −∑ a j φ j ≥0
i | j ⟩
 derive Bessel’s inequality ⟨ f |f ⟩ ≥ ∑ ¿ an ∨¿ ¿
2

Answer:
The φ j are assumed to be orthonormal. Expanding I, we have
I =⟨ f ∨f ⟩−∑ ai ⟨ φi∨f ⟩ −¿ ∑ ai ⟨ f ∨φi ⟩ + ∑ ai a j ⟨ φi∨φ j ⟩ ≥ 0 ¿
¿ ¿ ¿

i i i

Using the relation a i=⟨ φi∨f ⟩ and the orthonormality condition ⟨ φi ∨φ j ⟩

I =⟨ f ∨f ⟩−∑ ai ai −∑ ai ai + ∑ ai ai=⟨ f ∨f ⟩−∑ ¿ ai∨¿ ≥ 0¿
¿ ¿ ¿ 2

i i i i

5.1.8 Expand the function sinπx in a series of functions ϕi that are orthogonal (but not
normalized) on the range 0 ≤ x ≤ 1 when the scalar product has definition
1
⟨ f |g ⟩∫ f ¿ ( x ) g ( x ) dx
0

Keep the first four terms of the expansion. The first four ϕi are: ϕ0 = 1, ϕ1 = 2x − 1,
ϕ2 = 6x 2 − 6x + 1, ϕ3 = 20x 3 − 30x 2 + 12x − 1. Note. The integrals that are needed
are the subject of Example 1.10.5.
Answer:
The expansion we need is
⟨ φi ∨sin πx ⟩
sin πx=¿ ∑ φ (x)
i ⟨ φi∨φ i ⟩ i
The necessary integrals are
1 1
1
⟨ φ0 ∨φ0 ⟩=∫ dx=1 , ⟨ φ 1∨φ1 ⟩ =∫ ( 2 x−1 )2 dx= 3
0 0

1
1
⟨ φ2 ∨φ2 ⟩ =∫ ( 6 x −6 x +1 ) dx=5 ⟨ φ3 ∨φ3 ⟩ =
2 2

0
7
1 1
2
⟨ φ0 ∨f ⟩ =∫ sin πx dx= π
⟨ φ1∨f ⟩ =∫ ( 2 x−1 ) sin πx dx=0
0 0

2 24
⟨ φ2 ∨f ⟩= π − 3 ⟨ φ3 ∨f ⟩=0
π
 2 2 24
−
π π π3
φ2+ …=0.6366−0.6871 ( 6 x −6 x +1 ) …
2
sin πx= φ0 +
1 1
5
This series converges fairly rapidly. See Fig. 5.1.8.

5.1.9 Expand the function e −x in Laguerre polynomials Ln(x), which are orthonormal on the
range 0 ≤ x < ∞ with scalar product
∞
⟨ f |g ⟩∫ f ¿ ( x ) g ( x ) e−x dx
0

Keep the first four terms of the expansion. The first four Ln(x) are
2−4 x + x 2 6−18 x +9 x 2−x 3
L0=1 , L1=1−x , L2 = , L3 =
2 6
Answer:
−x
e =a0 L0 ( x ) +a 1 L1 ( x ) +a 2 L2 ( x ) +a 3 L3 ( x ) +… ,
∞
a 1=∫ Li (x) e
−2 x
dx
0

1 1 1 1
By integration we find a 0+ , a1= , a2= , a3= thus,
2 4 8 16

2 2 3
−x 1 1 1 2−4 x + x 1 6−18 x +9 x −x
e = ( 1 )+ ( 1−x ) + + +…,
2 4 8 2 16 6

This expansion when terminated after L3 fails badly beyond about x = 3. See Fig.
5.1.9.

5.1.10 The explicit form of a function f is not known, but the coefficients an of its expansion
in the orthonormal set ϕn are available. Assuming that the ϕn and the members of
another orthonormal set, χn , are available, use Dirac notation to obtain a formula for
the coefficients for the expansion of f in the χn set.
Answer:
 The forms ∑ ¿ φ i ⟩⟨ φi ∨¿ ¿ and ∑ ¿ φ j ⟩ ⟨ φ j ∨¿ ¿ are resolutions of the identity. Therefore
i j

¿ f ⟩ ∑ ¿ x j ⟩ ⟨ x j ¿ φi ⟩⟨ φi ¿ f ⟩ .
ij

The coefficients of f in the ϕ basis are ai = ⟨ φi ¿ f ⟩ , so the above equation is equivalent

to
f =∑ b j x j withb j ∑ ¿ ¿
j i

5.1.11 Using conventional vector notation, evaluate ∑ ¿ e^ j ⟩⟨ e^ j ¿ a ⟩, where a is an arbitrary

j

vector in the space spanned by the e^ j

Answer:
We assume the unit vectors are orthogonal. Then,
∑ ¿ e^ j ⟩⟨ e^ j ¿ a ⟩=∑ ( e^ j .a ) e^ j .
j i

This expression is a component decomposition of a.

5.1.12 Letting a=a1 e^ 1+ a2 e^ 2∧b=b1 e^ 1 +b2 e^ 2 be vectors in R2for what values of k ,if any, is
⟨ a1|b 1 ⟩ =a1 b1−a1 b 2−a2 b1 +ka 2 b 2
a valid definition of a scalar product?
Answer:
The scalar product ⟨ a|a ⟩ must be positive for every nonzero vector in the space. If we
2
write ha|ai in the form ( a 1−a2 )2 +(k – 1)a , this condition will be violated for some
2
nonzero a unless k > 1.
 ¿
5.2.1 Following the Gram-Schmidt procedure, construct a set of polynomials pn (x)
orthogonal (unit weighting factor) over the range [0, 1] from the set [1, x, x 2 , . . .].
Scale so that P ∗ n (1) = 1.

ANS. P ∗ n (x) = 1,

P ∗ 1 (x) = 2x − 1,

P ∗ 2 (x) = 6x 2 − 6x + 1,

P ∗ 3 (x) = 20x 3 − 30x 2 + 12x − 1.

These are the first four shifted Legendre polynomials. Note. The “*” is the standard
notation for “shifted”: [0, 1] instead of [−1, 1]. It does not mean complex conjugate.
Answer:
solution is given in the text. Note that
−1 −1 −1 −1 −9
a 10 = , a20= , a21 = , a 30= , a31= , a =−1/4 .
2 3 2 4 20 32

5.2.2 Apply the Gram-Schmidt procedure to form the first three Laguerre polynomials:
un ( x )=x n ,n=0,1,2 , … ,0 ≤ x <∞ , w ( x ) =e−x
Answer:
The solution is given in the text. Note that a 10=−1 , a20=−2 , a21=4

5.2.3 You are given

a) a set of functions un(x) =xn =0,1,2.....,

b) an interval (0. ꝏ)
c) a weighting function w(x)=xex. Use the Gram-Schmidt procedure to construct the
first three orthonormal functions from the set un.x/ for this interval and this
weighting function.
x−2 x 2−6 x +6
ϕ 0 ( x )=1 , ϕ1 ( x )= , ϕ 2 ( x )=
√2 2 √3

Answer:
The solution is given in the text. Note that a10 = −2, a20 = −6, a21 =−6 √ 2.

5.2.4 Using the Gram-Schmidt orthogonalization procedure, construct the lowest three
Hermite polynomials:
 2
n −x
un ( x )=x ,n=0,1,2 , … ,−ꝏ < x> ꝏ , w ( x )=e .
For this set of polynominals the usual normalization is

∫ hm ( x ) H n ( x ) ω ( x ) dx=⸹mn 2m m! π 1/ 2
−ꝏ

ANS. H0 = 1, H1 = 2x, H2 =4x2 -2

Answer:
The solution is given in the text. Note that a10 = 0, a20 = −1/2, a21 = 0

5.2.5 Use the Gram-Schmidt orthogonalization scheme to construct the first three
Chebyshev polynomials (type I):

un ( x )=x n ,n=0,1,2 , … ,−1 ≤ x ≤1 , ω ( x )=¿

Take the normalization

{
1 π ,∧m=n=0
∫ mT ( x ) ω ( x ) dx =⸹ mn π
,∧m=n ≥1
−1
2

Hint. The needed integrals are given in Exercise 13.3.2.

ANS. T0 = 1, T1 = x, T2 = 2x2-1, (T3=4x3-3x)

Answer:
Relying without comment on the integral formulas in Exercise 13.3.2, we compute
first

1
⟨ x 0 │ x0 ⟩=∫ ¿ ¿
−1

1
⟨x 1
│ x ⟩ = ⟨ x │ x ⟩= ∫ x ¿ ¿
1 0 2 2

−1

1
⟨ x 2 │ x 2 ⟩ =∫ x 4 ¿ ¿
−1

⟨ x 0 │ x1 ⟩ = ⟨ x 2 │ x1 ⟩=0

Note that some integrals are zero by symmetry

 The polynomial T0 is of the form c0 x0, with c0 satisfying

⟨ c 0 x 0 │ c 0 x 0 ⟩=│ c 0 │2 ⟨ x0 │ x 0 ⟩ =π
so c0 = 1 and T0 = 1. By symmetry, the polynomial T1, which in principle is a linear
combination of x0 and x1, must actually be an odd function that depends only on x1,
so is of the form c1x. It is automatically orthogonal to T0, and c1 must satisfy

π
⟨ c 1 x │c 1 x ⟩ =│ c 1 │2 ⟨ x │ x ⟩= 2

π
Because ⟨ x │ x ⟩ = , we have c 1=1 and T 1=x
2

The determination of T 1 is a bit less trivial. T 2 will be an even function of x, and will
be of the general form

⟨ T 2 │T 2 ⟩ =│ c2 │2 (⟨ x − 12 ) │ ( x − 12 )⟩
2 2

│ c2 │
2
( 38π − π2 + 14 )=│ c │ π8 = π2
2
2

from which we finnd c2 = 2 and T2 = 2x2 - 1.

5.2.6 Use the Gram-Schmidt orthogonalization scheme to construct the first three
Chebyshev

polynomials (type II):

n
un ( x )=x ,n=0,1,2 , … ,−1 ≤ x ≤1 ω ( x )=¿

Take the normalization to be

1 π
U m ( x ) U n ( x ) ω ( x ) dx=δ mn
∫ 2¿
−1
¿

Hint

∫¿¿
−1
 π
¿ , n=0
2

ANS U 0 =1, U 1=2 x ,U 2 =4 x 2−1

Answer:

From the formula given in the Hint, we have

⟨x 0
│ x 0 ⟩ =∫ ¿ ¿
−1

⟨ x 0 │ x2 ⟩ = ⟨ x 1 │ x1 ⟩= π
8

π
⟨ x2 │ x2 ⟩=
16

π
Taking U0 = c0x0, we find │ c 0 │ ⟨ x │ x 0 ⟩ = , so c0 = 1 and U0 = 1. The Un have
0
2
π π
even/odd symmetry, so U1 = c1x, and │ c 0 │ ⟨ x │ x 0 ⟩ = ,so │ c1 │ = and C1=2
0 2
2 8
U1=2x

finally

⟨ U 2 │ U 2 ⟩ =│ c1 │ 2
⟨( ) ( )⟩
1
4
2 2
x− │ x−
1
4
=
│ c 2 │2 π π
32
=
2

so c 2=4 , u2=4 x 2−1

5.2.7 As a modification of Exercise 5.2.5, apply the Gram-Schmidt orthogonalization

procedure to the setun ( x )=x n ,n=0,1,2 … , 0≤ x ≤ ꝏ . Take ω (x) to be exp(−x2 ). Find
the first two nonvanishing polynomials. Normalize so that the coefficient of
themhighest power of x is unity. In Exercise 5.2.5, the interval . (−ꝏ , ꝏ) led to the
Hermite polynomials. The functions found here are certainly not the Hermite
polynomials.

ANS φ 0=1 , φ1=x−π −1 /2

Answer:
The solution is given in the text. Note that a 10=−1/√ π
5.2.8 Form a set of three orthonormal vectors by the Gram-Schmidt process using these
input

vectors in the order given

1 1 1
c 1=1 c 2=1 c 3=0
1 2 2

Answer:
Let the orthonormalized vectors be denoted bi. First, Make b1 a nor- malized
version of c 1 : b1=c 1 /√ 3. Then obtain b2 (denoting b2 before normalization) as

() ( ) ( )
1 1 √3 −1 /3
b 2=c 2−( b 1 c 2 ) b1= 1 −( 4 √ 3 ) 1 √ 3 = −1 /3
2 1 √3 2 /3

Normalizing,b 2=√ 3/2b 2. Finally, form

( )(
−1
√6

() ( ) ( )
1 1√ 3
√3 −1 = 1/2
2 1√ 3
)
b 3=c 3−( b 1 c 3 ) b1 −( b2 c 3 ) b2= 0 −( √ 3 ) 1 √ 3 −
2 √6
−1/2
0
2
√6

()
−1
√6

( ) ( )
1 √3 1 / √2
−1
b 1= 1 √ 3 b 1= b 1= −1 / √2
√6
1 √3 0
2
√6

5.3.1 Show (without introducing matrix representations) that the adjoint of the adjoint
of an operator restores the original operator, i.e., that ¿
Answer:

⟨ φ │ Aψ ⟩=¿
 we may conclude that ¿

5.3.2 U and V are two arbitrary operators. Without introducing matrix representations
of these operators, show that

Note the resemblance to adjoint matrices.

Answer:

∫
(ψ2|V (U ψ1)) =
† †
(V ψ2)∗(U†ψ1)d3r

∫
= (UV ψ2)∗ψ1d3r = (ψ2|(UV )†ψ1).

5.3.3 Consider a Hilbert space spanned by the three functions φ 1=x 1 , φ2= x2 , φ3=x 3,
And a scalar product defined by ⟨ x v │ x μ ⟩ =δ vμ
a) Form the 3 x 3 matrix of each of the following operators

( ) ( )
3
∂ ∂ ∂
A1=∑ xi , A 2=x1 −x 2 ( )
i=1 ∂ xi ∂ x2 ∂ x1

Form the column vector representing ψ=x 1−2 x 2+ 3 x 3

b)

c) Form the matrix equation corresponding to χ =( A1 −A 2)ψ and verify that the
matrix equation reproduces the result obtained by direct application of A1− A 2
to ψ.
Answer:

( ) ( )
1 0 0 0 1 0
a). 1 0 1 0 ,
A = A2= −1 0 0
0 0 1 0 0 0
 ()
1
b). ψ= −2
3

( )( ) ( )
1 −1 0 1 3
c). ( A1 −A 2 ) ψ= 1 1 0 −2 = −1 =x
0 0 1 3 3

5.3.4

d
a) Obtain the matrix representation of A=x( ) in a basis of Legendre
dx
polynomials, keeping terms through P3. Use the orthonormal forms of these
polynomials as given in 5.2.1 and the scalar product defined there.
b) Expand x3 in the orthonormal Legendre polynomial basis
c) Verify that Ax3 is given correctly by its matrix representation
Answer:

a) Ap0=0 , Ap1=
√3 x= p ,
1
2

Ap2=
2 (2 2 )
√7 15 x 3− 3 x =3 p +√ 21 p ,
3 1

Using the above and noting that our basis is the Pn, we construct

( )
0 0 √5 0
A= 0 1 0 √ 21
0 0 2 0
0 0 0 3

( )
0
3
b) x →
√6 /5
0
2 √ 14 /35

c)

( )( ) ( )
0 0 √5 0 0 0
Ax = 0
3 1 0 √ 21 √ 6/5 = 3 √ 6/5
0 0 2 0 0 0
0 0 0 3 2 √14 /35 6 √14 /35

Inserting the explicit forms of P1 and P3, we find

 3
Ax = ( 3 √5 6 )( 2√3x )+( 6 √3514 ) √27 ( 52 x − 32 x)=3 x 3 3

in agreement with the directly computed value.

5.4.1

a) A is a non-Hermitian operator. Show that the operators A+ A+ ¿¿ and A−A +¿¿ are
hermitian.
b) Using the preceding result, show that every non-Hermitian operator may be written
as a linear combination of two Hermitian operators
Answer:

a). ¿

1
b). A= ¿
2

5.4.2

Prove that the product of two Hermitian operators is Hermitian if and only if the two
operators commute.

Answer:
+¿=BA =Aif only if [B , A] =0¿
+¿ A ¿

( AB )+¿=B ¿

5.4.3 A and B are noncommuting quantum mechanical operators, and C is given by the
formula

AB−BA=iC

Show that C is Hermitian. Assume that appropriate boundary conditions are satisfied.

Answer:
+ ¿=BA −AB=−iC ¿
+¿=−iC ¿
+¿ −A+¿ B ¿¿
+ ¿A ¿
+ ¿=B ¿
+¿=( iC ) ¿
( AB−BA )

5.4.4 The operator 𝓛 is Hermitian. Show that ⟨ L2 ⟩ ≥ 0, meaning that for all in the space in
 which 𝓛 is defined, ⟨ ψ │ L2 │ ψ ⟩ ≥ 0
+¿=BA =Aif only if [B , A] =0¿
+¿ A ¿

( AB )+¿=B ¿

5.4.5

Consider a Hilbert space whose members are functions defined on the surface of the
unit sphere, with a scalar product of the form

⟨ f │ g ⟩=∫ dΩ f ¿ g ,

where dΩ is the element of solid angle. Note that the total solid angle of the sphere is
4 π . We work here with the three functions φ 1=Cx/r , φ2 =Cy/r , φ3 =Cz/r , with C
asssigned a value that makes the φ i normalized.

a) Find C, and show that the φ i are also mutually orthogonal.

b) Form the 3 x 3 matrices of the angular momentum operator

(
L x =−i y
∂
∂z
−z
∂
∂y
, Ly =−i z
∂
∂x
−x
∂
∂z ) ( )
L z=−i x ( ∂
∂y
−y
∂
∂x )
5.5.1 Show that’s the matrix U of example 5.5.1 correctly transform the vector f(θ,φ)=
3x1+2ix2-x3+x5 to the { x '1} basis by:

(a). (1) making a column vector c that represent f(θ, φ) in the {x1} basis

(2) Forming c’ = Uc, and

(3) comparing the expansion ∑ c1 x 1 ( θ , φ ) ;

' '

(b) Verifying that U is unitary

Answer

()
3
2i
(a) (1) The column vector representing f (θ,φ) is c = −1
0
1
 ( )( ) ( )
−1/ √ 2 −i/√ 2 0 0 0 3 −1/ √ 2
1/ √ 2 −i/√ 2 0 0 0 2i 5 /√ 2
(2) c ' =¿ 0 0 i/√ 2 1/√ 2 0 −1 = = −i/√ 2
0 0 −i/√ 2 1/√ 2 0 0 i/√ 2
0 0 0 0 1 1 1

(3) Check: ∑ c ' i x ' i=∑ c i xi

(b) From U’ and verify that UU’=1

( )( )
−1/ √ 2 −i/√ 2 0 0 0 −1/√ 2 1/√ 2 0 0 0
1/ √ 2 −i/√ 2 0 0 0 i/√ 2 −i/√ 2 0 0 0
0 0 i/√ 2 1/√ 2 0 0 0 i/√ 2 i/√ 2 0 =1
0 0 −i/√ 2 1/√ 2 0 0 0 1/ √ 2 1/√ 2 0
0 0 0 0 1 0 0 0 0 1

5.5.2 (a) Given ( ¿ R3 ¿the basis φ 1=x , φ 2= y , φ3=z ,consider the basis transformation x
x → z , y → y , z−x , find the 3 ×3 matrix U for this transformation.

(b) This transformation corresponds to a rotation of the coordinate axes.

Identify the rotation and reconcile your transformation matrix with an appropriate
matrix S (α, β, γ) of the form given in Eq. (3.37).

(c) Form the column vector c representing (in the original basis) f = 2x − 3y + z, find
the result of applying U to c, and show that this is consistent with the basis
transformation of part (a).

Note. You do not need to be able to form scalar products to handle this exercise; a
knowledge of the linear relationship between the original and transformed functions
issufficient.

Answer

(a) the it column of U describes in the new basis, Thus

( )
0 0 1
U= 0 1 0
1 0 0
 (b) The transformation is counterlockwise rotation of the coordinate system about the
y axis; this correspon to the Euler angles are α=0, β= π/2, ϒ=0. The above U is
reproduced when theses angles are substituted into Eq. (3.37)

()
2
(c) c= −3
1

( )( ) ( )
0 0 −1 2 −1
Uc = 0 1 0 −3 = −3
1 0 0 1 2

This vector correspons to f’ = -x-3y+2z which is consistent with application of the

relevant basis transformation to f

5.5.3 Construct the matrix representing the inverse of the transformation in Exercise 5.5.2,
and show that this matrix and the transformation matrix of that exercise are matrix
inverses of each other.

Answer

Since the matrix U for transformation of the exercise 5.5.2 is unitary the inverse
transformation has matrix U, which is

( )
0 0 1
U= 0 1 0
−1 0 0

Multiplying, we find that Matrix UUt =1

5.5.4 The unitary transformation U that converts an orthonormal basis {φ i} into the basis {φ 'i
} and the unitary transformation V that converts the basis {φ 'i} into the basis {χi} have
matrix representations

( ) ( )
i sin θ cos θ 0 1 0 0
U = −cos θ i sin θ 0 , V = 0 cos θ isin θ
0 0 1 0 cos θ −isin θ

Given the function f(x) = 3φ 1 ( x )−φ 2 ( x )−2 φ3 ( x )

 (a) By applying U, form the vector representing f (x) in the {φ 'i } basis and then by
applying V form the vector representing f (x) in the {χi} basis. Use this result to
write f (x) as a linear combination of the χi.

(b) Form the matrix products UV and VU and then apply each to the vector
represent-ing f (x) in the { φ i} basis. Verify that the results of these applications
differ and that only one of them gives the result corresponding to part (a).

Answer:

( )( ) ( )
isin θ cos θ 0 3 cos θ+3 i sinθ
(a) Uf = −cos θ isin θ 0 −1 = −3 cos θ+i sin θ
0 0 1 −2 −2

( )( )
1 0 0 cos θ+3 i sinθ
V ( Uf ) = 0 cos θ i sinθ −3 cos θ+i sinθ
0 cos θ −i sinθ −2

¿¿

5.5.5 Three functions which are orthogonal with unit weight on the range −1 ≤ x ≤ 1 are P 0
2 2 1
= 1, P1 = x, and P2 = x − . Another set of functions that are orthogonal and span the
3 2
same space are F0 = x2, F1 = x, F2 = 5x2 −3. Although much of this exercise can be
done by inspection, write down and evaluate all the integrals that lead to the results
when they are obtained in terms of scalar products.

(a) Normalize each of the Pi and Fi

(b) Find the unitary matrix U that transforms from the normalized Pi basis to the
normalized Fi basis.

(c) Find the unitary matrix V that transforms from the normalized Fi basis to the
normalized Pi basis.

(d) Show that U and V are unitary, and that V = U−1

(e) Expand f (x) = 5x2 − 3x + 1 in terms of the normalized versions of both bases, and
verify that the transformation matrix U converts the P-basis expansion off (x) into its
F-basis expansion.
 Answer:

(a) The normalized versions of the Pn, denoted Pn are Pn=√(2 n+1)/2 Pn

The normalized versions of the Fn, denoted Pn are F 0 =√ 5/ 2 F 0

F1=√ 3/2 F1 and F2=√ 1/8 F2

(b) The transformation matrix U has elements uij=(Fi\Pj), for example:

1 1
√5 √5
u02 ∫ F0 P2 dx= ∫ 2
(x )(1¿) dx= ¿)
−1 2 −1 3

The complate transformation matrix is

( )
√5/3 0 2/3
U= 0 1 0
−2 /3 0 √5 /3

(c) V has elements vij = (Fi\Pj), thus

( )
√ 5/3 0 −2/3
V= 0 1 0
2/3 0 √5 /3

5.6.1 (a) Using the twi spin fanction φ 1=α , φ 2=β as and orthonormal basis (so ¿ α ∨α >¿) =
¿ β∨β >¿=1, ¿ α ∨β >¿=0) and the relations

1 1 1 −1 1 −1
S x α = β , S x β= α , S y α= iβ , S y β= iα , S z α = α , S z β= β
2 2 2 2 2 2

Construct the 2 ×2 matrices of S x , S y ∧S z

(b) Taking now the basis φ '1=C ( α + β ) ,=φ'2=C ( α −β ):

(i) Verify that φ '1and φ '2 are orthogonal
(ii) Assign C a Value that makes φ '1and φ '2 normalized
'
(iii) Find the unitary matrix for the transformation {φ 1 }→{φ 1 }

(c) Find the matrices of S x , S y ∧S z in the {φ '1 } basis

Answer
 (a) The first column of Sx shows the result of its operation on α; the second column
describes Sxβ. Similar observations apply to Sy amd Sz.We get

Sx= ( )
1 0 1
2 1 0
, Sy=
2 i 0(
1 0 −i
, Sz= )
1 1 0
2 0 −1
. ( )
(b) (1) Check that (α+β|α−β) = 0. Expanding, we have (α|αi− hα|βi+hβ|αi − hβ|βi = 1 +
0 + 0 − 1 = 0.

(2) A similar expansion shows that (α + β|α + β) = 2, so a propervalue of C is 1/√ 2.

The same result is obtained for ¿)

(3) The matrix elements of the transformation are uij = (φ ' 2 {φ ' ¿ ¿ 2)

These evaluate to

U= (1/1/ √√ 22 )
1/√ 2
= (
1 1 1
−1/ √ 2 √ 2 1 −1 )
(c) In the transford basis, the matrix of an operator S becomes S’ = U S U -1. Nothing
that U-1=U, we compute

s ' x =U x S U= ( )( )(
1 1 1 0 1 1 1
=
1 2 0
=
1 1 0
4 1 −1 1 0 1 −1 4 0 −2 2 1 −1
=s z ) ( ) ( )
Similar operation for s y ∧s z yield

s ' y =U S y U =
−1
(
1 0 i
2 −i 0
=−S y )
s ' z=U S z U =
−1
( )
1 0 1
2 1 0
=S z

2 2 2

5.6.2 For the basis φ 1=Cxe−r , φ 2=Cye−r , φ3 =Cze−r where r 2=x 2 + y 2 + z 2, with the scalar
product devined as an unweighted integral over R3 and with C chosen to make the φ i
normalized:

(a) Find the 3 ×3 matrix of L x =−i y ( ∂

∂z
−z
∂
∂y
; )
 ( )
1 0 0
(b) Using the transformations matrix U = 0 1/√ 2 −i/√ 2 find the transformed
0 1/√ 2 i/√ 2
matrix of Lx;

'
(c) Find the basis functions φ 1 defined by the transformation U, and write explicitly
(in term of x, y, z) the functional forms of L x φ'1 , i=1 ,2 , 3

Answer

(a) Apply L x to the φi; L xφ1=0 , L xφ2 =iφ 3 ,∧L xφ 3=−i φ2, therefore the matrix of Lx for
the basis is

( )
0 0 0
L x = 0 0 −i
0 i 0

(b) Form U Lx U−1. U is unitary; this can be checked by verifying that UU’=1 Thus,

( )( )( ) ( )
1 0 0 0 0 0 1 0 0 0 0 0
−1
U Lx U = 0 1/√ 2 −i / √ 2 0 0 −i 0 1/√ 2 1/√ 2 =¿ 0 1 0
0 1/√ 2 i/ √ 2 0 i 0 0 i /√ 2 −i/ √ 2 0 0 −1

(c) The new basis function have coefficients (in terms of the original basis) That are
the column of U’, reading them out, we have

−r 2 C 2
C 2

φ ' 1=Cxe , φ ' 2= ( y −iz ) e−r , φ ' 3= ( y −iz)e−r

√2 √2

Applying Lx,

Lφ ' 1= 0

C C C
( y−iz ) e−r = [ iz+i (−iy ) ] e−r = ( y −iz)e−r
2 2 2

Lφ ' 2=¿ Lx
√2 √2 √2
C C C
( y−iz ) e−r = [ iz−i (−iy ) ] e−r = ( y−iz) e−r
2 2 2

Lφ ' 3=¿
√2 √2 √2

5.6.3 The Gram-Schmidt process for converting an arbitrary basis X v into an orthonormal
set φν is described in Section 5.2 in a way that introduces coefficients of the form−
¿ φμ∨γv > ¿. For bases consisting of three functions, convert the formulation so that φ ν
 is expressed entirely in terms of the χµ, thereby obtaining an expression for the upper-
triangular matrix T appearing in Eq. (5.81).

Answer:

Define D1, D2, D3 as the determinants formed from the overlap matrix elements of
the first, the first two, and all three basis functions. Letting sij = [Xi |Xj] be the
elements of this overlap matrix, D 1 = S11, D2 = S11S22 − S12S21 etc. By substitution
into the formulas obtained as in Section 5.2, we find the systematic formulas for the
ϕi:

S 12 x1
X 2−
X1 D1
φ ' 1= , φ '2 =
√ D1 √ D1

S 12 X 1 D 1 S23 X 1 S12 S 23 X 1 S21 S 13 X 2 S 12 S21 S 13 X 1

X 3− − + + −
D1 D2 D2 D2 D1 D2
φ ' 1=
√ D3 / D2

Comparing with the matrix T as defined in Example 5.6.1, we see that its jth column
consists of the coefficients of the xi in the formula for ϕj. From the above formulas,
we find

( )
T 1 T 12 T 13
T = 0 T 22 T 23
0 0 T 33

With

1 S12 D
T 11 = , T 12= , T 22=√ 1
√ D1 √ D1 D2 D2

−D 2 S13 + D 1 S12 S 23−S 12 S21 S 13 −D 2 S 23+ S21 S 13

T 13 = ,T 23=
D1 √ D2 D3 √ D2 D 3

D2
T 23 =√
D3

5.7.1 Using the formal properties of unitary transformations, show that the commutator [x,
p] = i is invariant under unitary transformation of the matrices representing x and p
 Answer:

Replace x by x ' =U X U −1=¿ p ' =p by U p U−1, so

[ x ' , p' ¿=x ' p' − p' x ' =(U ¿¿ X U ¿¿−1)(U p U −1)−(U p U −1 )(U X U −1 ) ¿¿

−1 −1
= U (xp− px ) U =UU =i 1.

5.7.2 The Pauli matrices

σ 1= ( 01 10 ) , σ =(0i −i0 ) , σ =(10

2 3
0
−1 )
Have commutator [σ 1 , σ 2 ¿=2i σ 3 show that this relationship continues to be valid if
these matrices are transformed by

(
U = cos 2 θ sin 2θ
−sin2 θ cos 2θ )
Answer

We need

'
σ 1=U σ U =1
†
(cos
sin 2θ
2θ
cos 2θ
−sin2 θ
'
)
†
, σ 2=U σ U =
0 −i
i 0
, 2 ( )
'
σ 3=U σ U = 3
†
(−sin
cos 2 θ
2θ
−sin 2θ
−cos 2θ )
New form

' '
σ 1 σ 2= (−ii cossin22θθ −i sin 2θ
−i cos 2 θ
'
=iσ 3 )
' '
σ 2 σ 1= (−iisincos22θθ isin 2 θ
i cos 2 θ ) '
=−iσ 3

From the above, σ '1 σ '2−σ '2 σ '1=2i σ '3

5.7.3 (a) The operator L x is devined as

L x =−i y ( ∂
∂z
−z
∂
∂y )
 2 2 2

Verify that basis φ 1=Cxe−r , φ 2=Cye−r , φ3 =Cze−r ,where r 2=x 2 + y 2 + z 2, forms a

closed set under the operation of Lx, meaning when Lx applied to any member of this
basis the result is a function within the basis space , and construct the 3 ×3 matrix of
Lx in this basis from the result of the application of Lx to each basis function

(b) verify that L x [(x +iy)e r ]=−z e r and note that this result using the {φ i } the basis,
2 2

can be written L x ( φ 1+i φ2 ) =−φ3

(c) Express the equation of part (b) in matrix form, and write the matrix equation that
results when each of the quantities is transformed using the transformation matrix

( )
1 0 0
0 1/√ 2 −i/√ 2
0 1/√ 2 i/√ 2

(i) Regarding the transformation U as producing a new basis {φ' 1 } , find the explisit
form (in x, y, z) of the φ ' 1
(a) Using the operator of Lx and the explicit form of the φ ' 1, verify and validity of
the transformed equations found in part (c)

Answer

(a) From the equations Lxϕ1 = 0, Lxϕ2 = iϕ3, Lxϕ3 = −iϕ2, we see that Lx applied to any

function in the space spanned by ϕ1, ϕ2, ϕ3 a function that remains within that
space. The above equations correspond to the action on the ϕ basis of the
matrix

( )
0 0 0
L x = 0 0 −i
0 i 0

2
−r
(b) L x ( φ 1+iφ 2 )=0+i ( iφ 3 )=−φ3=−z e

( )( ) ( )
0 0 0 1 0
(c) 0 0 −i i = 0
0 i 0 0 −1

If this equations is transformed by U, the quantities in it become

 ( )( ) ( )( ) ( )
0 0 0 1 1 0 1
'
L x = 0 1 0 , i → i/√ 2 , 0 → i/√ 2
0 0 −1 0 i/√ 2 −1 −i/√ 2

And the transformed matrix equations is

( )( ) ( )
0 0 0 1 0
0 1 0 i /√ 2 = i/√ 2
0 0 −1 i /√ 2 −i/ √ 2

(d) The φ ' are those linear combinations of the φ with conficients that are
complex conjugates of the corresponding row of U and are

' −r
2 ( φ2 +iφ 3 ) −r
2
−r
2

φ 1=φ1=xe , φ 2= =( y−iz ) e /√ 2 ,φ 3=( φ 2+iφ 3 ) / √ 2 ( y −iz ) e /√ 2

√2

(e) the matrix equation is equivalent to

([ ( ) ( )] )
L x x+
i y +iz
√ 2 √2
+
i y−iz −r
√ 2 √2
e
2

¿
[ (
i y +iz
√2 √ 2
− )i y −iz −r
√2 √2
e ( )]
2

Which simplifies to

L x [(x +iy)e ] =−ze−r

2 2
−r

This is result that was proved in part (b).