SlideShare una empresa de Scribd logo
1 de 52
Descargar para leer sin conexión
1 / 29
Numerical Integration based on the Hyperfunction Theory
The 6th China-Japan-Korea Joint Conference on Numerical Mathematics
∗Hidenori Ogata (The University of Electro-Communications, Japan)
joint work with Hiroshi Hirayama (Kanaga Institute of Technology, Japan)
23 August, 2016
Contents
2 / 29
An application of the hyperfunction theory to numerical analysis
Contents
2 / 29
An application of the hyperfunction theory to numerical analysis
hyperfunction theory✓ ✏
generalized function theory proposed by M. Sato (1958)
based on the complex function theory
function with singularities
pole
discontinuity
delta function, ...
←−
complex analytic functions
familiar in
numerical computations
✒ ✑
Contents
2 / 29
An application of the hyperfunction theory to numerical analysis
hyperfunction theory✓ ✏
generalized function theory proposed by M. Sato (1958)
based on the complex function theory
function with singularities
pole
discontinuity
delta function, ...
←−
complex analytic functions
familiar in
numerical computations
✒ ✑
In this speech, we show a numerical integration method based on
the hyperfunction theory, which is expected to be efficient for integrals with
singularities.
Contents
3 / 29
1. Hyperfunction theory
2. Numerical integration over a finite interval
3. Numerical integration over an infinite interval
4. Numerical examples
5. Summary
Contents
4 / 29
1. Hyperfunction theory
2. Numerical integration over a finite interval
3. Numerical integration over an infinite interval
4. Numerical examples
5. Summary
1. Hyperfunction theory (an example)
5 / 29
b
a
φ(x)δ(x)dx = φ(0).
1
2πi C
φ(z)
z
dz = φ(0).
( a < 0 < b )
Dirac delta function Cauchy integral formula
O
C
D
φ(z) : analytic in D
1. Hyperfunction theory (an example)
5 / 29
b
a
φ(x)δ(x)dx = φ(0).
1
2πi C
φ(z)
z
dz = φ(0).
( a < 0 < b )
Dirac delta function Cauchy integral formula
O
C
D
Oa b
φ(z) : analytic in D
1
2πi C
φ(z)
z
dz = −
1
2πi
b
a
φ(x)
1
x + i0
−
1
x − i0
dx.
1. Hyperfunction theory (an example)
5 / 29
b
a
φ(x)δ(x)dx = φ(0).
1
2πi C
φ(z)
z
dz = φ(0).
( a < 0 < b )
Dirac delta function Cauchy integral formula
O
C
D
Oa b
φ(z) : analytic in D
b
a
φ(x)δ(x)dx =
1
2πi C
φ(z)
z
dz = −
1
2πi
b
a
φ(x)
1
x + i0
−
1
x − i0
dx.
∴ δ(x) = −
1
2πi
1
x + i0
−
1
x − i0
.
1. Hyperfunction theory
6 / 29
Hyperfunction theory (M. Sato, 1958)✓ ✏
• A hyperfunction f(x) on an interval I is the difference of the boundary
values of an analytic function F(z).
f(x) = [F(z)] ≡ F(x + i0) − F(x − i0).
F(z) : the defining function of the hyperfunction f(x)
analytic in D  I,
where D is a complex neighborhood of the interval I.
✒ ✑
D
I
R
1. Hyperfunction theory: examples
7 / 29
• Dirac delta function
δ(x) = −
1
2πiz
= −
1
2πi
1
x + i0
−
1
x − i0
.
• Heaviside step function
H(x) =
1 ( x > 0 )
0 ( x < 0 )
= −
1
2πi
{log(−(x + i0)) − log(−(x − i0))} .
∗ log z is the principal value s.t. −π ≦ arg z < π
• Non-integral powers
x+
α
=
xα ( x > 0 )
0 ( x < 0 )
= −
(−(x + i0))α − (−(x − i0))α
2i sin(πα)
(α ∈ Z const.).
∗ zα
is the principal value s.t. −π ≦ arg z < π.
1. Hyperfunction theory: examples
8 / 29
Heaviside step function
H(x) =
1 ( x > 0 )
0 ( x < 0 )
= F(x+i0)−F(x−i0), F(z) = −
1
2πi
log(−z).
-0.4 -0.2 0 0.2 0.4 0.6 0.8 1Re z
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
Im z
-1
-0.5
0
0.5
1
Re F(z)
The real part of the defining function F(z) = −
1
2πi
log(−z).
1. Hyperfunction theory: examples
8 / 29
Heaviside step function
H(x) =
1 ( x > 0 )
0 ( x < 0 )
= F(x+i0)−F(x−i0), F(z) = −
1
2πi
log(−z).
-0.4 -0.2 0 0.2 0.4 0.6 0.8 1Re z
-0.6
-0.4
-0.2
0
0.2
0.4
0.6
Im z
-1
-0.5
0
0.5
1
Re F(z)
Many functions with singularities are expressed by analytic functions
in the hyperfunction theory.
1. Hyperfunction theory: integral
9 / 29
Integral of a hyperfunction f(x) = F(x + i0) − F(x − i0)✓ ✏
I
f(x)dx ≡ −
C
F(z)dz
C : closed path which encircles I in the positive sense
and is included in D (F(z) is analytic in D  I).
✒ ✑
D
C
I
• The integral in independent of the choise of C by the Cauchy integral theorem.
1. Hyperfunction theory: integral
9 / 29
Integral of a hyperfunction f(x) = F(x + i0) − F(x − i0)✓ ✏
I
f(x)dx ≡ −
C
F(z)dz
C : closed path which encircles I in the positive sense
and is included in D (F(z) is analytic in D  I).
✒ ✑
D
I
I
f(x)dx =
I
{F(x + i0) − F(x − i0)} dx.
Contents
10 / 29
1. Hyperfunction theory
2. Numerical integration over a finite interval
3. Numerical integration over an infinite interval
4. Numerical examples
5. Summary
2. Numerical integration over a finite interval
11 / 29
We consider the evaluation of an integral
I
f(x)w(x)dx,
f(x) : analytic in a domain D s.t.
I ⊂ D ⊂ C,
w(x) : weight function.
D
I
R
2. Numerical integration over a finite interval
11 / 29
We consider the evaluation of an integral
I
f(x)w(x)dx,
f(x) : analytic in a domain D s.t.
I ⊂ D ⊂ C,
w(x) : weight function.
D
I
R
We regard the integrand as a hyperfunction.
✓ ✏
f(x)w(x)χI(x) = −
1
2πi
{f(x + i0)Ψ(x + i0) − f(x − i0)Ψ(x − i0)}
with χI(x) =
1 (x ∈ I)
0 (x ∈ I)
, Ψ(z) =
I
w(x)
z − x
dx
Hilbert
transform
✒ ✑
2. Numerical integration over a finite interval
12 / 29
From the definition of hyperfunction integrals, we have
✓ ✏
I
f(x)w(x)dx =
1
2πi C
f(z)Ψ(z)dz
=
1
2πi
uperiod
0
f(ϕ(u))Ψ(ϕ(u))ϕ′
(u)du,
C : z = ϕ(u) ( 0 ≦ u ≦ uperiod ) periodic function of period uperiod.
✒ ✑
D
C : z = ϕ(u)
I
Approximating the r.h.s. by the trapezoidal rule, we have ...
2. Numerical integration over a finite interval
13 / 29
Hyperfunction method✓ ✏
I
f(x)w(x)dx ≃
h
2πi
N−1
k=0
f(ϕ(kh))Ψ(ϕ(kh))ϕ′
(kh),
with Ψ(z) =
b
a
w(x)
z − x
dx and h =
uperiod
N
.
✒ ✑
D
C : z = ϕ(u), 0 ≦ u ≦ uperiod
I
2. Numerical integration over a finite interval
13 / 29
Hyperfunction method✓ ✏
I
f(x)w(x)dx ≃
h
2πi
N−1
k=0
f(ϕ(kh))Ψ(ϕ(kh))ϕ′
(kh),
with Ψ(z) =
b
a
w(x)
z − x
dx and h =
uperiod
N
.
✒ ✑
Ψ(z) for typical weight functions w(x)
I w(x) Ψ(z)
(a, b) 1 log
z − a
z − b
∗
(0, 1) xα−1(1 − x)β−1 B(α, β)z−1F(α, 1; α + β; z−1)∗∗
( α, β > 0 )
∗ log z is the principal value s.t. −π ≦ arg z < π.
∗∗ F(α, 1; α + β; z−1
) can be easily evaluated by the continued fraction.
2. Numerical integration over a finite interval
13 / 29
Hyperfunction method✓ ✏
I
f(x)w(x)dx ≃
h
2πi
N−1
k=0
f(ϕ(kh))Ψ(ϕ(kh))ϕ′
(kh),
with Ψ(z) =
b
a
w(x)
z − x
dx and h =
uperiod
N
.
✒ ✑
In the hyperfunction method, we numerically evaluate the complex integral
which defines the desired integral as a hyperfunction integral.
2. Numerical integration over a finite interval
14 / 29
If f(x) is real valued for real x, we can reduce the number of sampling points
N by half by the reflection principle.
I
f(x)w(x)dx ≃
h
π
Im
1
2
f(ϕ(0))Ψ(ϕ(0))ϕ′
(0) +
1
2
f(ϕ(Nh))Ψ(ϕ(Nh))ϕ′
(Nh)
+
N′
k=1
f(ϕ(kh))Ψ(ϕ(kh))ϕ′
(kh) ,
where z = ϕ(u) is a parameterization of C s.t.
ϕ(−u) = ϕ(u) and h =
uperiod
2N′
.
D
C : z = ϕ(u)
I
2. Numerical integration over a finite interval (error estimate)
15 / 29
The hyperfunction method is very accurate.
∵ The trapezoidal rule is efficient for the integrals of periodic analytic functions.
2. Numerical integration over a finite interval (error estimate)
15 / 29
The hyperfunction method is very accurate.
∵ The trapezoidal rule is efficient for the integrals of periodic analytic functions.
Theoretical error estimate✓ ✏
If f(ϕ(w)) and ϕ(w) are analytic in | Im w| < d0, the error of
the hyperfunction method (with N reduction by half) is bounded
by the inequality
|error| ≦ 2uperiod max
Im w=±d
|f(ϕ(w))Ψ(ϕ(w))ϕ′
(w)|
×
exp(−(4πd/uperiod)N)
1 − exp(−(4πd/uperiod)N)
( 0 < ∀d < d0 ).
. . . Geometric convergence.
✒ ✑
Contents
16 / 29
1. Hyperfunction theory
2. Numerical integration over a finite interval
3. Numerical integration over an infinite interval
4. Numerical examples
5. Summary
3. Numerical integration over an infinite interval
17 / 29
We consider the evaluation of an integral over the infinite interval (0, +∞)
∞
0
f(x)w(x)dx,
f(x) : analytic in a domain D s.t.
[0, +∞) ⊂ D ⊂ C,
w(x) : weight function.
D
R
O
3. Numerical integration over an infinite interval
17 / 29
We consider the evaluation of an integral over the infinite interval (0, +∞)
∞
0
f(x)w(x)dx,
f(x) : analytic in a domain D s.t.
[0, +∞) ⊂ D ⊂ C,
w(x) : weight function.
D
R
O
We regard the integrand as a hyperfunction.
✓ ✏
f(x)w(x)H(x) = −
1
2πi
{f(x + i0)Ψ(x + i0) − f(x − i0)Ψ(x − i0)}
with H(x) =
1 (x > 0)
0 (x < 0)
(Heaviside’s step function),
w(x) xα−1 ( α > 0, α ∈ Z ) 1
Ψ(z) −πzα−1/ sin(πα) − log(−z)
✒ ✑
3. Numerical integration over an infinite interval
18 / 29
From the definition of hyperfunction integrals, we have
✓ ✏
∞
0
f(x)w(x)dx =
1
2πi C
f(z)Ψ(z)dz
=
1
2πi
+∞
−∞
f(ϕ(u))Ψ(ϕ(u))ϕ′
(u)du,
C : z = ϕ(u) ( −∞ < u < +∞ ).
✒ ✑
R
D
C : z = ϕ(u)
O
Approximating the r.h.s. by the trapezoidal rule, we have ...
3. Numerical integration over an infinite interval
19 / 29
Hyperfunction method✓ ✏
∞
0
f(x)w(x)dx ≃
h
2πi
∞
k=−∞
f(ϕ(kh))Ψ(ϕ(kh))ϕ′
(kh)
≃
h
2πi
N1
k=−N2
f(ϕ(kh))Ψ(ϕ(kh))ϕ′
(kh).
✒ ✑
R
D
C : z = ϕ(u)
O
∗ Actually, we use the DE transform
to make the convergence of the infinite sum fast.
3. Numerical integration over an infinite interval
19 / 29
Hyperfunction method✓ ✏
∞
0
f(x)w(x)dx ≃
h
2πi
∞
k=−∞
f(ϕ(kh))Ψ(ϕ(kh))ϕ′
(kh)
≃
h
2πi
N1
k=−N2
f(ϕ(kh))Ψ(ϕ(kh))ϕ′
(kh).
✒ ✑
R
D
C : z = ϕ(u)
O
∗ We can reduce the number of sampling points N by half
by the reflection principle also in the cases of infinite intervals.
Contents
20 / 29
1. Hyperfunction theory
2. Numerical integration over a finite interval
3. Numerical integration over an infinite interval
4. Numerical examples
5. Summary
4. Example 1: numerical integration over a finite interval
21 / 29
1
0
ex
xα−1
(1 − x)β−1
dx = B(α, β)F(α; α + β; 1) ( α, β > 0 ).
We evaluated the integral by
• the hyperfunction method (with N reduction by half)
• the DE formula (efficient for integrals with end-point singularities)
and compared the errors of the two methods.
• C++ programs, double precision.
• integral path for the hyperfunction method
z =
1
2
+
1
4
ρ +
1
ρ
cos u +
i
4
ρ −
1
ρ
sin u ( ρ = 10 )
= 0.5 + 2.575 cos u + i2.425 sin u.
4. Example 1: numerical integration over a finite interval
22 / 29
-16
-14
-12
-10
-8
-6
-4
-2
0
0 10 20 30 40 50 60
log10(relativeerror)
N
hyperfunction rule
DE rule
-16
-14
-12
-10
-8
-6
-4
-2
0
0 10 20 30 40 50 60
log10(relativeerror)
N
hyperfunction rule
DE rule
α = 0.5 α = 10−4 (very strong singularity)
The errors of the hyperfunction method and the DE rule.
hyperfunction method DE rule
α = 0.5 O(0.025N ) O(0.36N )
α = 10−4 O(0.025N ) —
4. Example 1: numerical integration over a finite interval
22 / 29
-16
-14
-12
-10
-8
-6
-4
-2
0
0 10 20 30 40 50 60
log10(relativeerror)
N
hyperfunction rule
DE rule
-16
-14
-12
-10
-8
-6
-4
-2
0
0 10 20 30 40 50 60
log10(relativeerror)
N
hyperfunction rule
DE rule
α = 0.5 α = 10−4 (very strong singularity)
The errors of the hyperfunction method and the DE rule.
hyperfunction method DE rule
α = 0.5 O(0.025N ) O(0.36N )
α = 10−4 O(0.025N ) —
The DE rule does not work if the end-point singularities are very strong.
4. Example 1: numerical integration over a finite interval
22 / 29
-16
-14
-12
-10
-8
-6
-4
-2
0
0 10 20 30 40 50 60
log10(relativeerror)
N
hyperfunction rule
DE rule
-16
-14
-12
-10
-8
-6
-4
-2
0
0 10 20 30 40 50 60
log10(relativeerror)
N
hyperfunction rule
DE rule
α = 0.5 α = 10−4 (very strong singularity)
The errors of the hyperfunction method and the DE rule.
hyperfunction method DE rule
α = 0.5 O(0.025N ) O(0.36N )
α = 10−4 O(0.025N ) —
The convergence of the hyperfunction method is not affected
by the end-point singularities.
4. Example 2: numerical integration over a finite interval
23 / 29
1
0
xα−1(1 − x)β−1
1 + x2
dx = B(α, β) Re{F(α, 1; α + β; i)} ( α, β > 0 ).
• integral path for the hyperfunction method
z =
1
2
+
1
4
ρ +
1
ρ
cos u +
i
4
ρ −
1
ρ
sin u ( ρ = 2 )
= 0.5 + 0.625 cos u + i0.375 sin u.
4. Example 2: numerical integration over a finite interval
23 / 29
1
0
xα−1(1 − x)β−1
1 + x2
dx = B(α, β) Re{F(α, 1; α + β; i)} ( α, β > 0 ).
-16
-14
-12
-10
-8
-6
-4
-2
0
0 10 20 30 40 50 60
log10(relativeerror)
N
hyperfunction method
DE rule
-16
-14
-12
-10
-8
-6
-4
-2
0
0 10 20 30 40 50 60
log10(relativeerror)
N
hyperfunction method
DE rule
α = 0.5 α = 10−4 (very strong singularity)
The errors of the hyperfunction method and the DE rule.
4. Example 2: numerical integration over a finite interval
23 / 29
1
0
xα−1(1 − x)β−1
1 + x2
dx = B(α, β) Re{F(α, 1; α + β; i)} ( α, β > 0 ).
The errors of the hyperfunction method and the DE rule.
hyperfunction method DE rule
α = 0.5 O(0.18N ) O(0.54N )
α = 10−4 O(0.25N ) —
• The DE rule does not work if the end-point singularities are very strong.
• The convergence of the hyperfunction method is not affected
by the end-point singularities.
4. Example: Why the hyperfunction method works well?
24 / 29
integrand
e z
hyperfunction method
• (DE rule) The sampling points accumulate at the singularities.
• (hyperfunction method) The sampling points are distributed on a curve
in the complex plane where the integrand varies slowly.
4. Example: Why the hyperfunction method works well?
24 / 29
integrand
e z
hyperfunction method
• (DE rule) The sampling points accumulate at the singularities.
• (hyperfunction method) The sampling points are distributed on a curve
in the complex plane where the integrand varies slowly.
Thus, the hyperfunction method is not affected by the end-point
singularities.
4. Example 3: Numerical integration over an infinite interval
25 / 29
∞
0
xα−1
1 + x2
dx =
π/2
sin(πα/2)
We computed it by the hyperfunction method (with N reduction by half)
and the DE rule.
• C++ program & double precision
• integral path
z = ϕ(u) =
w(u)
iπ
log
1 + iw(u)
1 − iw(u)
,
w = sinh(sinh u)
DE transform
+0.5i.
-3
-2
-1
0
1
2
3
-3 -2 -1 0 1 2 3
Imz Re z
4. Example 3: Numerical integration over an infinite interval
26 / 29
-16
-14
-12
-10
-8
-6
-4
-2
0
0 20 40 60 80 100
log10(relativeerror)
N
hyperfunction method
DE rule
-16
-14
-12
-10
-8
-6
-4
-2
0
0 20 40 60 80 100
log10(relativeerror)
N
hyperfunction method
DE rule
α = 0.5 α = 10−4 (very strong singularity)
The error of the hyperfunction method and the DE rule.
hyperfunction method DE rule
α = 0.5 O(0.51N ) O(0.34N )
α = 10−4 O(0.46N ) O(0.57N )
4. Example 3: Numerical integration over an infinite interval
26 / 29
-16
-14
-12
-10
-8
-6
-4
-2
0
0 20 40 60 80 100
log10(relativeerror)
N
hyperfunction method
DE rule
-16
-14
-12
-10
-8
-6
-4
-2
0
0 20 40 60 80 100
log10(relativeerror)
N
hyperfunction method
DE rule
α = 0.5 α = 10−4 (very strong singularity)
The error of the hyperfunction method and the DE rule.
hyperfunction method DE rule
α = 0.5 O(0.51N ) O(0.34N )
α = 10−4 O(0.46N ) O(0.57N )
The convergence rate of the hyperfunction method is not affected
by the end-point singularity.
4. Example 4: Numerical integration over an infinite interval
27 / 29
∞
0
xα−1
e−x
dx = Γ(α) ( α > 0 ).
• integral path for the hyperfunction
method
z = ϕ(u) =
w(u)
iπ
log
1 + iw(u)
1 − iw(u)
,
w = sinh u
DE transform
+0.5i.
-3
-2
-1
0
1
2
3
-3 -2 -1 0 1 2 3
Imz
Re z
4. Example 4: Numerical integration over an infinite interval
27 / 29
∞
0
xα−1
e−x
dx = Γ(α) ( α > 0 ).
-16
-14
-12
-10
-8
-6
-4
-2
0
0 20 40 60 80 100
log10(error)
N
alpha=0.5
alpha=0.1
alpha=0.01
alpha=1.0e-4
-16
-14
-12
-10
-8
-6
-4
-2
0
0 20 40 60 80 100
log10(error)
N
d=0.5
d=0.1
d=0.01
d=1.0e-4
hyperfunction method DE rule
The erros of the hyperfunction method and the DE rule.
4. Example 4: Numerical integration over an infinite interval
27 / 29
∞
0
xα−1
e−x
dx = Γ(α) ( α > 0 ).
The errors of the hyperfunction method and the DE rule.
α 0.5 0.1 0.01 10−4
hyperfunction rule O(0.46N ) O(0.46N ) O(0.46N ) O(0.46N )
error
DE rule O(0.39N ) O(0.47N ) O(0.53N ) O(0.55N )
The convergence of the hyperfunction method is not affected
by the end-point singularities.
Contents
28 / 29
1. Hyperfunction theory
2. Numerical integration over a finite interval
3. Numerical integration over an infinite interval
4. Numerical examples
5. Summary
5. Summary
29 / 29
• The hyperfunction theory is a generalized function theory based on the
complex function theory.
• The hyperfunction method approximately computes desired integral by
evaluating the complex integrals which define them as hyperfunction
integrals
• Numerical examples show that the hyperfunction method is efficient for
integral with end-point singularities.
5. Summary
29 / 29
• The hyperfunction theory is a generalized function theory based on the
complex function theory.
• The hyperfunction method approximately computes desired integral by
evaluating the complex integrals which define them as hyperfunction
integrals
• Numerical examples show that the hyperfunction method is efficient for
integral with end-point singularities.
functions with singularities
(poles, discontinuities,
delta functions, ...)
←−←−←−
hyperfunction
analytic
functions
5. Summary
29 / 29
• The hyperfunction theory is a generalized function theory based on the
complex function theory.
• The hyperfunction method approximately computes desired integral by
evaluating the complex integrals which define them as hyperfunction
integrals
• Numerical examples show that the hyperfunction method is efficient for
integral with end-point singularities.
functions with singularities
(poles, discontinuities,
delta functions, ...)
←−←−←−
hyperfunction
analytic
functions
Hyperfunctions connect singular functions with analytic functions.
We expect that we can apply the hyperfunction theory to a wide range of
scientific computations.
5. Summary
29 / 29
• The hyperfunction theory is a generalized function theory based on the
complex function theory.
• The hyperfunction method approximately computes desired integral by
evaluating the complex integrals which define them as hyperfunction
integrals
• Numerical examples show that the hyperfunction method is efficient for
integral with end-point singularities.
functions with singularities
(poles, discontinuities,
delta functions, ...)
←−←−←−
hyperfunction
analytic
functions
Hyperfunctions connect singular functions with analytic functions.
We expect that we can apply the hyperfunction theory to a wide range of
scientific computations.
Thank you!

Más contenido relacionado

La actualidad más candente

Bayesian hybrid variable selection under generalized linear models
Bayesian hybrid variable selection under generalized linear modelsBayesian hybrid variable selection under generalized linear models
Bayesian hybrid variable selection under generalized linear models
Caleb (Shiqiang) Jin
 

La actualidad más candente (20)

Numerical differentiation integration
Numerical differentiation integrationNumerical differentiation integration
Numerical differentiation integration
 
Inequality, slides #2
Inequality, slides #2Inequality, slides #2
Inequality, slides #2
 
A series of maximum entropy upper bounds of the differential entropy
A series of maximum entropy upper bounds of the differential entropyA series of maximum entropy upper bounds of the differential entropy
A series of maximum entropy upper bounds of the differential entropy
 
QMC Program: Trends and Advances in Monte Carlo Sampling Algorithms Workshop,...
QMC Program: Trends and Advances in Monte Carlo Sampling Algorithms Workshop,...QMC Program: Trends and Advances in Monte Carlo Sampling Algorithms Workshop,...
QMC Program: Trends and Advances in Monte Carlo Sampling Algorithms Workshop,...
 
Brief Introduction About Topological Interference Management (TIM)
Brief Introduction About Topological Interference Management (TIM)Brief Introduction About Topological Interference Management (TIM)
Brief Introduction About Topological Interference Management (TIM)
 
Intro to Approximate Bayesian Computation (ABC)
Intro to Approximate Bayesian Computation (ABC)Intro to Approximate Bayesian Computation (ABC)
Intro to Approximate Bayesian Computation (ABC)
 
ABC with data cloning for MLE in state space models
ABC with data cloning for MLE in state space modelsABC with data cloning for MLE in state space models
ABC with data cloning for MLE in state space models
 
Intro to ABC
Intro to ABCIntro to ABC
Intro to ABC
 
Newton's Backward Interpolation Formula with Example
Newton's Backward Interpolation Formula with ExampleNewton's Backward Interpolation Formula with Example
Newton's Backward Interpolation Formula with Example
 
From moments to sparse representations, a geometric, algebraic and algorithmi...
From moments to sparse representations, a geometric, algebraic and algorithmi...From moments to sparse representations, a geometric, algebraic and algorithmi...
From moments to sparse representations, a geometric, algebraic and algorithmi...
 
Accelerated approximate Bayesian computation with applications to protein fol...
Accelerated approximate Bayesian computation with applications to protein fol...Accelerated approximate Bayesian computation with applications to protein fol...
Accelerated approximate Bayesian computation with applications to protein fol...
 
QMC Program: Trends and Advances in Monte Carlo Sampling Algorithms Workshop,...
QMC Program: Trends and Advances in Monte Carlo Sampling Algorithms Workshop,...QMC Program: Trends and Advances in Monte Carlo Sampling Algorithms Workshop,...
QMC Program: Trends and Advances in Monte Carlo Sampling Algorithms Workshop,...
 
Distributed Architecture of Subspace Clustering and Related
Distributed Architecture of Subspace Clustering and RelatedDistributed Architecture of Subspace Clustering and Related
Distributed Architecture of Subspace Clustering and Related
 
Interpolation In Numerical Methods.
 Interpolation In Numerical Methods. Interpolation In Numerical Methods.
Interpolation In Numerical Methods.
 
Multiattribute utility copula
Multiattribute utility copulaMultiattribute utility copula
Multiattribute utility copula
 
Probability Formula sheet
Probability Formula sheetProbability Formula sheet
Probability Formula sheet
 
Madrid easy
Madrid easyMadrid easy
Madrid easy
 
MUMS Opening Workshop - An Overview of Reduced-Order Models and Emulators (ED...
MUMS Opening Workshop - An Overview of Reduced-Order Models and Emulators (ED...MUMS Opening Workshop - An Overview of Reduced-Order Models and Emulators (ED...
MUMS Opening Workshop - An Overview of Reduced-Order Models and Emulators (ED...
 
Bayesian hybrid variable selection under generalized linear models
Bayesian hybrid variable selection under generalized linear modelsBayesian hybrid variable selection under generalized linear models
Bayesian hybrid variable selection under generalized linear models
 
QMC Program: Trends and Advances in Monte Carlo Sampling Algorithms Workshop,...
QMC Program: Trends and Advances in Monte Carlo Sampling Algorithms Workshop,...QMC Program: Trends and Advances in Monte Carlo Sampling Algorithms Workshop,...
QMC Program: Trends and Advances in Monte Carlo Sampling Algorithms Workshop,...
 

Similar a Numerical integration based on the hyperfunction theory

3130005 cvpde gtu_study_material_e-notes_all_18072019070728_am
3130005 cvpde gtu_study_material_e-notes_all_18072019070728_am3130005 cvpde gtu_study_material_e-notes_all_18072019070728_am
3130005 cvpde gtu_study_material_e-notes_all_18072019070728_am
dataniyaarunkumar
 
Indefinite Integral
Indefinite IntegralIndefinite Integral
Indefinite Integral
JelaiAujero
 

Similar a Numerical integration based on the hyperfunction theory (20)

Maximum likelihood estimation of regularisation parameters in inverse problem...
Maximum likelihood estimation of regularisation parameters in inverse problem...Maximum likelihood estimation of regularisation parameters in inverse problem...
Maximum likelihood estimation of regularisation parameters in inverse problem...
 
Tensor Train data format for uncertainty quantification
Tensor Train data format for uncertainty quantificationTensor Train data format for uncertainty quantification
Tensor Train data format for uncertainty quantification
 
Maths AIP.pdf
Maths AIP.pdfMaths AIP.pdf
Maths AIP.pdf
 
IVR - Chapter 1 - Introduction
IVR - Chapter 1 - IntroductionIVR - Chapter 1 - Introduction
IVR - Chapter 1 - Introduction
 
Differential Calculus
Differential Calculus Differential Calculus
Differential Calculus
 
Distributed solution of stochastic optimal control problem on GPUs
Distributed solution of stochastic optimal control problem on GPUsDistributed solution of stochastic optimal control problem on GPUs
Distributed solution of stochastic optimal control problem on GPUs
 
Spectral sum rules for conformal field theories
Spectral sum rules for conformal field theoriesSpectral sum rules for conformal field theories
Spectral sum rules for conformal field theories
 
SOLVING BVPs OF SINGULARLY PERTURBED DISCRETE SYSTEMS
SOLVING BVPs OF SINGULARLY PERTURBED DISCRETE SYSTEMSSOLVING BVPs OF SINGULARLY PERTURBED DISCRETE SYSTEMS
SOLVING BVPs OF SINGULARLY PERTURBED DISCRETE SYSTEMS
 
1519 differentiation-integration-02
1519 differentiation-integration-021519 differentiation-integration-02
1519 differentiation-integration-02
 
1531 fourier series- integrals and trans
1531 fourier series- integrals and trans1531 fourier series- integrals and trans
1531 fourier series- integrals and trans
 
The integral
The integralThe integral
The integral
 
3130005 cvpde gtu_study_material_e-notes_all_18072019070728_am
3130005 cvpde gtu_study_material_e-notes_all_18072019070728_am3130005 cvpde gtu_study_material_e-notes_all_18072019070728_am
3130005 cvpde gtu_study_material_e-notes_all_18072019070728_am
 
Litvinenko_RWTH_UQ_Seminar_talk.pdf
Litvinenko_RWTH_UQ_Seminar_talk.pdfLitvinenko_RWTH_UQ_Seminar_talk.pdf
Litvinenko_RWTH_UQ_Seminar_talk.pdf
 
2.1 Calculus 2.formulas.pdf.pdf
2.1 Calculus 2.formulas.pdf.pdf2.1 Calculus 2.formulas.pdf.pdf
2.1 Calculus 2.formulas.pdf.pdf
 
Murphy: Machine learning A probabilistic perspective: Ch.9
Murphy: Machine learning A probabilistic perspective: Ch.9Murphy: Machine learning A probabilistic perspective: Ch.9
Murphy: Machine learning A probabilistic perspective: Ch.9
 
Stochastic Frank-Wolfe for Constrained Finite Sum Minimization @ Montreal Opt...
Stochastic Frank-Wolfe for Constrained Finite Sum Minimization @ Montreal Opt...Stochastic Frank-Wolfe for Constrained Finite Sum Minimization @ Montreal Opt...
Stochastic Frank-Wolfe for Constrained Finite Sum Minimization @ Montreal Opt...
 
Hierarchical matrices for approximating large covariance matries and computin...
Hierarchical matrices for approximating large covariance matries and computin...Hierarchical matrices for approximating large covariance matries and computin...
Hierarchical matrices for approximating large covariance matries and computin...
 
Colloquium
ColloquiumColloquium
Colloquium
 
Indefinite Integral
Indefinite IntegralIndefinite Integral
Indefinite Integral
 
Linear response theory and TDDFT
Linear response theory and TDDFT Linear response theory and TDDFT
Linear response theory and TDDFT
 

Último

Call Girls in Ramesh Nagar Delhi 💯 Call Us 🔝9953056974 🔝 Escort Service
Call Girls in Ramesh Nagar Delhi 💯 Call Us 🔝9953056974 🔝 Escort ServiceCall Girls in Ramesh Nagar Delhi 💯 Call Us 🔝9953056974 🔝 Escort Service
Call Girls in Ramesh Nagar Delhi 💯 Call Us 🔝9953056974 🔝 Escort Service
9953056974 Low Rate Call Girls In Saket, Delhi NCR
 
Structural Analysis and Design of Foundations: A Comprehensive Handbook for S...
Structural Analysis and Design of Foundations: A Comprehensive Handbook for S...Structural Analysis and Design of Foundations: A Comprehensive Handbook for S...
Structural Analysis and Design of Foundations: A Comprehensive Handbook for S...
Dr.Costas Sachpazis
 
result management system report for college project
result management system report for college projectresult management system report for college project
result management system report for college project
Tonystark477637
 
UNIT-V FMM.HYDRAULIC TURBINE - Construction and working
UNIT-V FMM.HYDRAULIC TURBINE - Construction and workingUNIT-V FMM.HYDRAULIC TURBINE - Construction and working
UNIT-V FMM.HYDRAULIC TURBINE - Construction and working
rknatarajan
 

Último (20)

(INDIRA) Call Girl Aurangabad Call Now 8617697112 Aurangabad Escorts 24x7
(INDIRA) Call Girl Aurangabad Call Now 8617697112 Aurangabad Escorts 24x7(INDIRA) Call Girl Aurangabad Call Now 8617697112 Aurangabad Escorts 24x7
(INDIRA) Call Girl Aurangabad Call Now 8617697112 Aurangabad Escorts 24x7
 
University management System project report..pdf
University management System project report..pdfUniversity management System project report..pdf
University management System project report..pdf
 
Call Girls Pimpri Chinchwad Call Me 7737669865 Budget Friendly No Advance Boo...
Call Girls Pimpri Chinchwad Call Me 7737669865 Budget Friendly No Advance Boo...Call Girls Pimpri Chinchwad Call Me 7737669865 Budget Friendly No Advance Boo...
Call Girls Pimpri Chinchwad Call Me 7737669865 Budget Friendly No Advance Boo...
 
VIP Model Call Girls Kothrud ( Pune ) Call ON 8005736733 Starting From 5K to ...
VIP Model Call Girls Kothrud ( Pune ) Call ON 8005736733 Starting From 5K to ...VIP Model Call Girls Kothrud ( Pune ) Call ON 8005736733 Starting From 5K to ...
VIP Model Call Girls Kothrud ( Pune ) Call ON 8005736733 Starting From 5K to ...
 
Call Girls in Ramesh Nagar Delhi 💯 Call Us 🔝9953056974 🔝 Escort Service
Call Girls in Ramesh Nagar Delhi 💯 Call Us 🔝9953056974 🔝 Escort ServiceCall Girls in Ramesh Nagar Delhi 💯 Call Us 🔝9953056974 🔝 Escort Service
Call Girls in Ramesh Nagar Delhi 💯 Call Us 🔝9953056974 🔝 Escort Service
 
Coefficient of Thermal Expansion and their Importance.pptx
Coefficient of Thermal Expansion and their Importance.pptxCoefficient of Thermal Expansion and their Importance.pptx
Coefficient of Thermal Expansion and their Importance.pptx
 
Booking open Available Pune Call Girls Koregaon Park 6297143586 Call Hot Ind...
Booking open Available Pune Call Girls Koregaon Park  6297143586 Call Hot Ind...Booking open Available Pune Call Girls Koregaon Park  6297143586 Call Hot Ind...
Booking open Available Pune Call Girls Koregaon Park 6297143586 Call Hot Ind...
 
PVC VS. FIBERGLASS (FRP) GRAVITY SEWER - UNI BELL
PVC VS. FIBERGLASS (FRP) GRAVITY SEWER - UNI BELLPVC VS. FIBERGLASS (FRP) GRAVITY SEWER - UNI BELL
PVC VS. FIBERGLASS (FRP) GRAVITY SEWER - UNI BELL
 
Structural Analysis and Design of Foundations: A Comprehensive Handbook for S...
Structural Analysis and Design of Foundations: A Comprehensive Handbook for S...Structural Analysis and Design of Foundations: A Comprehensive Handbook for S...
Structural Analysis and Design of Foundations: A Comprehensive Handbook for S...
 
UNIT-II FMM-Flow Through Circular Conduits
UNIT-II FMM-Flow Through Circular ConduitsUNIT-II FMM-Flow Through Circular Conduits
UNIT-II FMM-Flow Through Circular Conduits
 
result management system report for college project
result management system report for college projectresult management system report for college project
result management system report for college project
 
Vivazz, Mieres Social Housing Design Spain
Vivazz, Mieres Social Housing Design SpainVivazz, Mieres Social Housing Design Spain
Vivazz, Mieres Social Housing Design Spain
 
Intze Overhead Water Tank Design by Working Stress - IS Method.pdf
Intze Overhead Water Tank  Design by Working Stress - IS Method.pdfIntze Overhead Water Tank  Design by Working Stress - IS Method.pdf
Intze Overhead Water Tank Design by Working Stress - IS Method.pdf
 
Booking open Available Pune Call Girls Pargaon 6297143586 Call Hot Indian Gi...
Booking open Available Pune Call Girls Pargaon  6297143586 Call Hot Indian Gi...Booking open Available Pune Call Girls Pargaon  6297143586 Call Hot Indian Gi...
Booking open Available Pune Call Girls Pargaon 6297143586 Call Hot Indian Gi...
 
Thermal Engineering Unit - I & II . ppt
Thermal Engineering  Unit - I & II . pptThermal Engineering  Unit - I & II . ppt
Thermal Engineering Unit - I & II . ppt
 
Call for Papers - International Journal of Intelligent Systems and Applicatio...
Call for Papers - International Journal of Intelligent Systems and Applicatio...Call for Papers - International Journal of Intelligent Systems and Applicatio...
Call for Papers - International Journal of Intelligent Systems and Applicatio...
 
UNIT-V FMM.HYDRAULIC TURBINE - Construction and working
UNIT-V FMM.HYDRAULIC TURBINE - Construction and workingUNIT-V FMM.HYDRAULIC TURBINE - Construction and working
UNIT-V FMM.HYDRAULIC TURBINE - Construction and working
 
(INDIRA) Call Girl Bhosari Call Now 8617697112 Bhosari Escorts 24x7
(INDIRA) Call Girl Bhosari Call Now 8617697112 Bhosari Escorts 24x7(INDIRA) Call Girl Bhosari Call Now 8617697112 Bhosari Escorts 24x7
(INDIRA) Call Girl Bhosari Call Now 8617697112 Bhosari Escorts 24x7
 
The Most Attractive Pune Call Girls Budhwar Peth 8250192130 Will You Miss Thi...
The Most Attractive Pune Call Girls Budhwar Peth 8250192130 Will You Miss Thi...The Most Attractive Pune Call Girls Budhwar Peth 8250192130 Will You Miss Thi...
The Most Attractive Pune Call Girls Budhwar Peth 8250192130 Will You Miss Thi...
 
data_management_and _data_science_cheat_sheet.pdf
data_management_and _data_science_cheat_sheet.pdfdata_management_and _data_science_cheat_sheet.pdf
data_management_and _data_science_cheat_sheet.pdf
 

Numerical integration based on the hyperfunction theory

  • 1. 1 / 29 Numerical Integration based on the Hyperfunction Theory The 6th China-Japan-Korea Joint Conference on Numerical Mathematics ∗Hidenori Ogata (The University of Electro-Communications, Japan) joint work with Hiroshi Hirayama (Kanaga Institute of Technology, Japan) 23 August, 2016
  • 2. Contents 2 / 29 An application of the hyperfunction theory to numerical analysis
  • 3. Contents 2 / 29 An application of the hyperfunction theory to numerical analysis hyperfunction theory✓ ✏ generalized function theory proposed by M. Sato (1958) based on the complex function theory function with singularities pole discontinuity delta function, ... ←− complex analytic functions familiar in numerical computations ✒ ✑
  • 4. Contents 2 / 29 An application of the hyperfunction theory to numerical analysis hyperfunction theory✓ ✏ generalized function theory proposed by M. Sato (1958) based on the complex function theory function with singularities pole discontinuity delta function, ... ←− complex analytic functions familiar in numerical computations ✒ ✑ In this speech, we show a numerical integration method based on the hyperfunction theory, which is expected to be efficient for integrals with singularities.
  • 5. Contents 3 / 29 1. Hyperfunction theory 2. Numerical integration over a finite interval 3. Numerical integration over an infinite interval 4. Numerical examples 5. Summary
  • 6. Contents 4 / 29 1. Hyperfunction theory 2. Numerical integration over a finite interval 3. Numerical integration over an infinite interval 4. Numerical examples 5. Summary
  • 7. 1. Hyperfunction theory (an example) 5 / 29 b a φ(x)δ(x)dx = φ(0). 1 2πi C φ(z) z dz = φ(0). ( a < 0 < b ) Dirac delta function Cauchy integral formula O C D φ(z) : analytic in D
  • 8. 1. Hyperfunction theory (an example) 5 / 29 b a φ(x)δ(x)dx = φ(0). 1 2πi C φ(z) z dz = φ(0). ( a < 0 < b ) Dirac delta function Cauchy integral formula O C D Oa b φ(z) : analytic in D 1 2πi C φ(z) z dz = − 1 2πi b a φ(x) 1 x + i0 − 1 x − i0 dx.
  • 9. 1. Hyperfunction theory (an example) 5 / 29 b a φ(x)δ(x)dx = φ(0). 1 2πi C φ(z) z dz = φ(0). ( a < 0 < b ) Dirac delta function Cauchy integral formula O C D Oa b φ(z) : analytic in D b a φ(x)δ(x)dx = 1 2πi C φ(z) z dz = − 1 2πi b a φ(x) 1 x + i0 − 1 x − i0 dx. ∴ δ(x) = − 1 2πi 1 x + i0 − 1 x − i0 .
  • 10. 1. Hyperfunction theory 6 / 29 Hyperfunction theory (M. Sato, 1958)✓ ✏ • A hyperfunction f(x) on an interval I is the difference of the boundary values of an analytic function F(z). f(x) = [F(z)] ≡ F(x + i0) − F(x − i0). F(z) : the defining function of the hyperfunction f(x) analytic in D I, where D is a complex neighborhood of the interval I. ✒ ✑ D I R
  • 11. 1. Hyperfunction theory: examples 7 / 29 • Dirac delta function δ(x) = − 1 2πiz = − 1 2πi 1 x + i0 − 1 x − i0 . • Heaviside step function H(x) = 1 ( x > 0 ) 0 ( x < 0 ) = − 1 2πi {log(−(x + i0)) − log(−(x − i0))} . ∗ log z is the principal value s.t. −π ≦ arg z < π • Non-integral powers x+ α = xα ( x > 0 ) 0 ( x < 0 ) = − (−(x + i0))α − (−(x − i0))α 2i sin(πα) (α ∈ Z const.). ∗ zα is the principal value s.t. −π ≦ arg z < π.
  • 12. 1. Hyperfunction theory: examples 8 / 29 Heaviside step function H(x) = 1 ( x > 0 ) 0 ( x < 0 ) = F(x+i0)−F(x−i0), F(z) = − 1 2πi log(−z). -0.4 -0.2 0 0.2 0.4 0.6 0.8 1Re z -0.6 -0.4 -0.2 0 0.2 0.4 0.6 Im z -1 -0.5 0 0.5 1 Re F(z) The real part of the defining function F(z) = − 1 2πi log(−z).
  • 13. 1. Hyperfunction theory: examples 8 / 29 Heaviside step function H(x) = 1 ( x > 0 ) 0 ( x < 0 ) = F(x+i0)−F(x−i0), F(z) = − 1 2πi log(−z). -0.4 -0.2 0 0.2 0.4 0.6 0.8 1Re z -0.6 -0.4 -0.2 0 0.2 0.4 0.6 Im z -1 -0.5 0 0.5 1 Re F(z) Many functions with singularities are expressed by analytic functions in the hyperfunction theory.
  • 14. 1. Hyperfunction theory: integral 9 / 29 Integral of a hyperfunction f(x) = F(x + i0) − F(x − i0)✓ ✏ I f(x)dx ≡ − C F(z)dz C : closed path which encircles I in the positive sense and is included in D (F(z) is analytic in D I). ✒ ✑ D C I • The integral in independent of the choise of C by the Cauchy integral theorem.
  • 15. 1. Hyperfunction theory: integral 9 / 29 Integral of a hyperfunction f(x) = F(x + i0) − F(x − i0)✓ ✏ I f(x)dx ≡ − C F(z)dz C : closed path which encircles I in the positive sense and is included in D (F(z) is analytic in D I). ✒ ✑ D I I f(x)dx = I {F(x + i0) − F(x − i0)} dx.
  • 16. Contents 10 / 29 1. Hyperfunction theory 2. Numerical integration over a finite interval 3. Numerical integration over an infinite interval 4. Numerical examples 5. Summary
  • 17. 2. Numerical integration over a finite interval 11 / 29 We consider the evaluation of an integral I f(x)w(x)dx, f(x) : analytic in a domain D s.t. I ⊂ D ⊂ C, w(x) : weight function. D I R
  • 18. 2. Numerical integration over a finite interval 11 / 29 We consider the evaluation of an integral I f(x)w(x)dx, f(x) : analytic in a domain D s.t. I ⊂ D ⊂ C, w(x) : weight function. D I R We regard the integrand as a hyperfunction. ✓ ✏ f(x)w(x)χI(x) = − 1 2πi {f(x + i0)Ψ(x + i0) − f(x − i0)Ψ(x − i0)} with χI(x) = 1 (x ∈ I) 0 (x ∈ I) , Ψ(z) = I w(x) z − x dx Hilbert transform ✒ ✑
  • 19. 2. Numerical integration over a finite interval 12 / 29 From the definition of hyperfunction integrals, we have ✓ ✏ I f(x)w(x)dx = 1 2πi C f(z)Ψ(z)dz = 1 2πi uperiod 0 f(ϕ(u))Ψ(ϕ(u))ϕ′ (u)du, C : z = ϕ(u) ( 0 ≦ u ≦ uperiod ) periodic function of period uperiod. ✒ ✑ D C : z = ϕ(u) I Approximating the r.h.s. by the trapezoidal rule, we have ...
  • 20. 2. Numerical integration over a finite interval 13 / 29 Hyperfunction method✓ ✏ I f(x)w(x)dx ≃ h 2πi N−1 k=0 f(ϕ(kh))Ψ(ϕ(kh))ϕ′ (kh), with Ψ(z) = b a w(x) z − x dx and h = uperiod N . ✒ ✑ D C : z = ϕ(u), 0 ≦ u ≦ uperiod I
  • 21. 2. Numerical integration over a finite interval 13 / 29 Hyperfunction method✓ ✏ I f(x)w(x)dx ≃ h 2πi N−1 k=0 f(ϕ(kh))Ψ(ϕ(kh))ϕ′ (kh), with Ψ(z) = b a w(x) z − x dx and h = uperiod N . ✒ ✑ Ψ(z) for typical weight functions w(x) I w(x) Ψ(z) (a, b) 1 log z − a z − b ∗ (0, 1) xα−1(1 − x)β−1 B(α, β)z−1F(α, 1; α + β; z−1)∗∗ ( α, β > 0 ) ∗ log z is the principal value s.t. −π ≦ arg z < π. ∗∗ F(α, 1; α + β; z−1 ) can be easily evaluated by the continued fraction.
  • 22. 2. Numerical integration over a finite interval 13 / 29 Hyperfunction method✓ ✏ I f(x)w(x)dx ≃ h 2πi N−1 k=0 f(ϕ(kh))Ψ(ϕ(kh))ϕ′ (kh), with Ψ(z) = b a w(x) z − x dx and h = uperiod N . ✒ ✑ In the hyperfunction method, we numerically evaluate the complex integral which defines the desired integral as a hyperfunction integral.
  • 23. 2. Numerical integration over a finite interval 14 / 29 If f(x) is real valued for real x, we can reduce the number of sampling points N by half by the reflection principle. I f(x)w(x)dx ≃ h π Im 1 2 f(ϕ(0))Ψ(ϕ(0))ϕ′ (0) + 1 2 f(ϕ(Nh))Ψ(ϕ(Nh))ϕ′ (Nh) + N′ k=1 f(ϕ(kh))Ψ(ϕ(kh))ϕ′ (kh) , where z = ϕ(u) is a parameterization of C s.t. ϕ(−u) = ϕ(u) and h = uperiod 2N′ . D C : z = ϕ(u) I
  • 24. 2. Numerical integration over a finite interval (error estimate) 15 / 29 The hyperfunction method is very accurate. ∵ The trapezoidal rule is efficient for the integrals of periodic analytic functions.
  • 25. 2. Numerical integration over a finite interval (error estimate) 15 / 29 The hyperfunction method is very accurate. ∵ The trapezoidal rule is efficient for the integrals of periodic analytic functions. Theoretical error estimate✓ ✏ If f(ϕ(w)) and ϕ(w) are analytic in | Im w| < d0, the error of the hyperfunction method (with N reduction by half) is bounded by the inequality |error| ≦ 2uperiod max Im w=±d |f(ϕ(w))Ψ(ϕ(w))ϕ′ (w)| × exp(−(4πd/uperiod)N) 1 − exp(−(4πd/uperiod)N) ( 0 < ∀d < d0 ). . . . Geometric convergence. ✒ ✑
  • 26. Contents 16 / 29 1. Hyperfunction theory 2. Numerical integration over a finite interval 3. Numerical integration over an infinite interval 4. Numerical examples 5. Summary
  • 27. 3. Numerical integration over an infinite interval 17 / 29 We consider the evaluation of an integral over the infinite interval (0, +∞) ∞ 0 f(x)w(x)dx, f(x) : analytic in a domain D s.t. [0, +∞) ⊂ D ⊂ C, w(x) : weight function. D R O
  • 28. 3. Numerical integration over an infinite interval 17 / 29 We consider the evaluation of an integral over the infinite interval (0, +∞) ∞ 0 f(x)w(x)dx, f(x) : analytic in a domain D s.t. [0, +∞) ⊂ D ⊂ C, w(x) : weight function. D R O We regard the integrand as a hyperfunction. ✓ ✏ f(x)w(x)H(x) = − 1 2πi {f(x + i0)Ψ(x + i0) − f(x − i0)Ψ(x − i0)} with H(x) = 1 (x > 0) 0 (x < 0) (Heaviside’s step function), w(x) xα−1 ( α > 0, α ∈ Z ) 1 Ψ(z) −πzα−1/ sin(πα) − log(−z) ✒ ✑
  • 29. 3. Numerical integration over an infinite interval 18 / 29 From the definition of hyperfunction integrals, we have ✓ ✏ ∞ 0 f(x)w(x)dx = 1 2πi C f(z)Ψ(z)dz = 1 2πi +∞ −∞ f(ϕ(u))Ψ(ϕ(u))ϕ′ (u)du, C : z = ϕ(u) ( −∞ < u < +∞ ). ✒ ✑ R D C : z = ϕ(u) O Approximating the r.h.s. by the trapezoidal rule, we have ...
  • 30. 3. Numerical integration over an infinite interval 19 / 29 Hyperfunction method✓ ✏ ∞ 0 f(x)w(x)dx ≃ h 2πi ∞ k=−∞ f(ϕ(kh))Ψ(ϕ(kh))ϕ′ (kh) ≃ h 2πi N1 k=−N2 f(ϕ(kh))Ψ(ϕ(kh))ϕ′ (kh). ✒ ✑ R D C : z = ϕ(u) O ∗ Actually, we use the DE transform to make the convergence of the infinite sum fast.
  • 31. 3. Numerical integration over an infinite interval 19 / 29 Hyperfunction method✓ ✏ ∞ 0 f(x)w(x)dx ≃ h 2πi ∞ k=−∞ f(ϕ(kh))Ψ(ϕ(kh))ϕ′ (kh) ≃ h 2πi N1 k=−N2 f(ϕ(kh))Ψ(ϕ(kh))ϕ′ (kh). ✒ ✑ R D C : z = ϕ(u) O ∗ We can reduce the number of sampling points N by half by the reflection principle also in the cases of infinite intervals.
  • 32. Contents 20 / 29 1. Hyperfunction theory 2. Numerical integration over a finite interval 3. Numerical integration over an infinite interval 4. Numerical examples 5. Summary
  • 33. 4. Example 1: numerical integration over a finite interval 21 / 29 1 0 ex xα−1 (1 − x)β−1 dx = B(α, β)F(α; α + β; 1) ( α, β > 0 ). We evaluated the integral by • the hyperfunction method (with N reduction by half) • the DE formula (efficient for integrals with end-point singularities) and compared the errors of the two methods. • C++ programs, double precision. • integral path for the hyperfunction method z = 1 2 + 1 4 ρ + 1 ρ cos u + i 4 ρ − 1 ρ sin u ( ρ = 10 ) = 0.5 + 2.575 cos u + i2.425 sin u.
  • 34. 4. Example 1: numerical integration over a finite interval 22 / 29 -16 -14 -12 -10 -8 -6 -4 -2 0 0 10 20 30 40 50 60 log10(relativeerror) N hyperfunction rule DE rule -16 -14 -12 -10 -8 -6 -4 -2 0 0 10 20 30 40 50 60 log10(relativeerror) N hyperfunction rule DE rule α = 0.5 α = 10−4 (very strong singularity) The errors of the hyperfunction method and the DE rule. hyperfunction method DE rule α = 0.5 O(0.025N ) O(0.36N ) α = 10−4 O(0.025N ) —
  • 35. 4. Example 1: numerical integration over a finite interval 22 / 29 -16 -14 -12 -10 -8 -6 -4 -2 0 0 10 20 30 40 50 60 log10(relativeerror) N hyperfunction rule DE rule -16 -14 -12 -10 -8 -6 -4 -2 0 0 10 20 30 40 50 60 log10(relativeerror) N hyperfunction rule DE rule α = 0.5 α = 10−4 (very strong singularity) The errors of the hyperfunction method and the DE rule. hyperfunction method DE rule α = 0.5 O(0.025N ) O(0.36N ) α = 10−4 O(0.025N ) — The DE rule does not work if the end-point singularities are very strong.
  • 36. 4. Example 1: numerical integration over a finite interval 22 / 29 -16 -14 -12 -10 -8 -6 -4 -2 0 0 10 20 30 40 50 60 log10(relativeerror) N hyperfunction rule DE rule -16 -14 -12 -10 -8 -6 -4 -2 0 0 10 20 30 40 50 60 log10(relativeerror) N hyperfunction rule DE rule α = 0.5 α = 10−4 (very strong singularity) The errors of the hyperfunction method and the DE rule. hyperfunction method DE rule α = 0.5 O(0.025N ) O(0.36N ) α = 10−4 O(0.025N ) — The convergence of the hyperfunction method is not affected by the end-point singularities.
  • 37. 4. Example 2: numerical integration over a finite interval 23 / 29 1 0 xα−1(1 − x)β−1 1 + x2 dx = B(α, β) Re{F(α, 1; α + β; i)} ( α, β > 0 ). • integral path for the hyperfunction method z = 1 2 + 1 4 ρ + 1 ρ cos u + i 4 ρ − 1 ρ sin u ( ρ = 2 ) = 0.5 + 0.625 cos u + i0.375 sin u.
  • 38. 4. Example 2: numerical integration over a finite interval 23 / 29 1 0 xα−1(1 − x)β−1 1 + x2 dx = B(α, β) Re{F(α, 1; α + β; i)} ( α, β > 0 ). -16 -14 -12 -10 -8 -6 -4 -2 0 0 10 20 30 40 50 60 log10(relativeerror) N hyperfunction method DE rule -16 -14 -12 -10 -8 -6 -4 -2 0 0 10 20 30 40 50 60 log10(relativeerror) N hyperfunction method DE rule α = 0.5 α = 10−4 (very strong singularity) The errors of the hyperfunction method and the DE rule.
  • 39. 4. Example 2: numerical integration over a finite interval 23 / 29 1 0 xα−1(1 − x)β−1 1 + x2 dx = B(α, β) Re{F(α, 1; α + β; i)} ( α, β > 0 ). The errors of the hyperfunction method and the DE rule. hyperfunction method DE rule α = 0.5 O(0.18N ) O(0.54N ) α = 10−4 O(0.25N ) — • The DE rule does not work if the end-point singularities are very strong. • The convergence of the hyperfunction method is not affected by the end-point singularities.
  • 40. 4. Example: Why the hyperfunction method works well? 24 / 29 integrand e z hyperfunction method • (DE rule) The sampling points accumulate at the singularities. • (hyperfunction method) The sampling points are distributed on a curve in the complex plane where the integrand varies slowly.
  • 41. 4. Example: Why the hyperfunction method works well? 24 / 29 integrand e z hyperfunction method • (DE rule) The sampling points accumulate at the singularities. • (hyperfunction method) The sampling points are distributed on a curve in the complex plane where the integrand varies slowly. Thus, the hyperfunction method is not affected by the end-point singularities.
  • 42. 4. Example 3: Numerical integration over an infinite interval 25 / 29 ∞ 0 xα−1 1 + x2 dx = π/2 sin(πα/2) We computed it by the hyperfunction method (with N reduction by half) and the DE rule. • C++ program & double precision • integral path z = ϕ(u) = w(u) iπ log 1 + iw(u) 1 − iw(u) , w = sinh(sinh u) DE transform +0.5i. -3 -2 -1 0 1 2 3 -3 -2 -1 0 1 2 3 Imz Re z
  • 43. 4. Example 3: Numerical integration over an infinite interval 26 / 29 -16 -14 -12 -10 -8 -6 -4 -2 0 0 20 40 60 80 100 log10(relativeerror) N hyperfunction method DE rule -16 -14 -12 -10 -8 -6 -4 -2 0 0 20 40 60 80 100 log10(relativeerror) N hyperfunction method DE rule α = 0.5 α = 10−4 (very strong singularity) The error of the hyperfunction method and the DE rule. hyperfunction method DE rule α = 0.5 O(0.51N ) O(0.34N ) α = 10−4 O(0.46N ) O(0.57N )
  • 44. 4. Example 3: Numerical integration over an infinite interval 26 / 29 -16 -14 -12 -10 -8 -6 -4 -2 0 0 20 40 60 80 100 log10(relativeerror) N hyperfunction method DE rule -16 -14 -12 -10 -8 -6 -4 -2 0 0 20 40 60 80 100 log10(relativeerror) N hyperfunction method DE rule α = 0.5 α = 10−4 (very strong singularity) The error of the hyperfunction method and the DE rule. hyperfunction method DE rule α = 0.5 O(0.51N ) O(0.34N ) α = 10−4 O(0.46N ) O(0.57N ) The convergence rate of the hyperfunction method is not affected by the end-point singularity.
  • 45. 4. Example 4: Numerical integration over an infinite interval 27 / 29 ∞ 0 xα−1 e−x dx = Γ(α) ( α > 0 ). • integral path for the hyperfunction method z = ϕ(u) = w(u) iπ log 1 + iw(u) 1 − iw(u) , w = sinh u DE transform +0.5i. -3 -2 -1 0 1 2 3 -3 -2 -1 0 1 2 3 Imz Re z
  • 46. 4. Example 4: Numerical integration over an infinite interval 27 / 29 ∞ 0 xα−1 e−x dx = Γ(α) ( α > 0 ). -16 -14 -12 -10 -8 -6 -4 -2 0 0 20 40 60 80 100 log10(error) N alpha=0.5 alpha=0.1 alpha=0.01 alpha=1.0e-4 -16 -14 -12 -10 -8 -6 -4 -2 0 0 20 40 60 80 100 log10(error) N d=0.5 d=0.1 d=0.01 d=1.0e-4 hyperfunction method DE rule The erros of the hyperfunction method and the DE rule.
  • 47. 4. Example 4: Numerical integration over an infinite interval 27 / 29 ∞ 0 xα−1 e−x dx = Γ(α) ( α > 0 ). The errors of the hyperfunction method and the DE rule. α 0.5 0.1 0.01 10−4 hyperfunction rule O(0.46N ) O(0.46N ) O(0.46N ) O(0.46N ) error DE rule O(0.39N ) O(0.47N ) O(0.53N ) O(0.55N ) The convergence of the hyperfunction method is not affected by the end-point singularities.
  • 48. Contents 28 / 29 1. Hyperfunction theory 2. Numerical integration over a finite interval 3. Numerical integration over an infinite interval 4. Numerical examples 5. Summary
  • 49. 5. Summary 29 / 29 • The hyperfunction theory is a generalized function theory based on the complex function theory. • The hyperfunction method approximately computes desired integral by evaluating the complex integrals which define them as hyperfunction integrals • Numerical examples show that the hyperfunction method is efficient for integral with end-point singularities.
  • 50. 5. Summary 29 / 29 • The hyperfunction theory is a generalized function theory based on the complex function theory. • The hyperfunction method approximately computes desired integral by evaluating the complex integrals which define them as hyperfunction integrals • Numerical examples show that the hyperfunction method is efficient for integral with end-point singularities. functions with singularities (poles, discontinuities, delta functions, ...) ←−←−←− hyperfunction analytic functions
  • 51. 5. Summary 29 / 29 • The hyperfunction theory is a generalized function theory based on the complex function theory. • The hyperfunction method approximately computes desired integral by evaluating the complex integrals which define them as hyperfunction integrals • Numerical examples show that the hyperfunction method is efficient for integral with end-point singularities. functions with singularities (poles, discontinuities, delta functions, ...) ←−←−←− hyperfunction analytic functions Hyperfunctions connect singular functions with analytic functions. We expect that we can apply the hyperfunction theory to a wide range of scientific computations.
  • 52. 5. Summary 29 / 29 • The hyperfunction theory is a generalized function theory based on the complex function theory. • The hyperfunction method approximately computes desired integral by evaluating the complex integrals which define them as hyperfunction integrals • Numerical examples show that the hyperfunction method is efficient for integral with end-point singularities. functions with singularities (poles, discontinuities, delta functions, ...) ←−←−←− hyperfunction analytic functions Hyperfunctions connect singular functions with analytic functions. We expect that we can apply the hyperfunction theory to a wide range of scientific computations. Thank you!