faculteit Wiskunde en

Natuurwetenschappen

An Introduction to
Differintegrals and
Fractional Calculus

Bachelor Research Mathematics

June 2013

Student: M. Jeeninga

Primary Supervisor: A.J. van der Schaft

Secondary Supervisor: H.S.V. de Snoo
 Abstract
We attempt to introduce differintegrals and fractional calculus to un-
dergraduates. After a brief history of fractional calculus and some prelim-
inary definitions, the concept of differintegrals is introduced: an operator
connecting differentiation and integration. Next we look at the Grünwald-
Letnikov and Riemann-Liouville differintegrals, what their construction is
based on and what connects them. We will be checking whether some of
the regular rules for differentiation still apply for the Riemann-Liouville
differintegral and conclude by calculating fractional integrals and frac-
tional derivatives of some basic functions.

2
 An Introduction to Differintegrals and
Fractional Calculus
Mark Jeeninga
s1961497
June 24, 2013

1 Introduction
In this paper we attempt to introduce differintegrals and fractional calculus to
undergraduates. After a brief history of fractional calculus and some basic def-
initions we will introduce the concept of differintegrals. Next we look at the
Grünwald-Letnikov and Riemann-Liouville differintegrals and see what their
construction is based on and what connects them. We will check whether some
regular rules for differentiation still apply for the Riemann-Liouville differinte-
gral. We conclude by calculating fractional integrals and fractional derivatives
of some basic functions.

1.1 History of Fractional Calculus

Leibniz invented infinitesimal calculus around 1674, independent from Newton.
In 1695 Leibniz wondered whether his calculus was compatible with non-integer
order derivatives, giving birth to the concept of fractional calculus. When sug-
gesting his question to l’Hôpital in a letter, L’Hôpital replied “What if the order
will be 21 ?”, referring to order of the derivative. Leibniz replied: “It leads to a
paradox, from which one day useful consequences will be drawn”.

Though the nature of this paradox is unclear, a relatively small group of mathe-
maticians started working on extending the regular methods of integration and
differentiation for non-integer orders. Fractional calculus started to mature over
the last 300 years. Over the course of time many different definitions for frac-
tional derivatives and integrals arose from different perspectives. Some even
seemed inconsistent with each other.

As Nishimoto wrote, fractional calculus is the calculus of the 21st century. It

slowly began to find applications in- and outside of mathematics, like in electri-
cal circuits and fluid mechanics. One particular case was solving the problem for
tautochrone motion using fractional calculus, found by Swedish mathematician
Abel in the 1820s ([3] p.31-35).

3
1.2 Preliminary Definitions
This paper will concern two common formulas for integration and differentia-
tion. We will be extending these formulas to non-integer order. We will address
these two formulas in a moment. The first one will be the backward difference
derivative and gives rise to the Grünwald-Letnikov differintegral. The second
one is the Cauchy formula for repeated integration and gives rise to the Riemann-
Liouville differintegral.

It is crucial to extend the factorial to non-integer values, in order to extend

the two formulas properly. That’s why we will also introduce the Gamma func-
tion, along with the Beta function.

1.2.1 Forward and Backward Difference

The forward difference of a function in a point t is given by

4h f (t) = f (t + h) − f (t)

Likewise, the backward difference is given by

5h f (t) = f (t) − f (t − h)

From this we can construct the forward difference derivative (FDD) and the
backward difference derivative (BDD):

d 4h f (t) f (t + h) − f (t)
(FDD) dt f (t) = lim = lim (1)
h→0 h h→0 h
d 5h f (t) f (t) − f (t − h)
(BDD) dt f (t) = h→0
lim = lim (2)
h h→0 h
Applying the FDD n times to a function f (t) we find
n  
X n
4nh f (t) = 4h 4h · · · 4h f (t) = (−1)k f (t + (n − k)h)
| {z } k
n k=0

For the BDD we find

n  
X n
5nh f (t) = (−1)k f (t − kh)
k
k=0

We can express the n-th derivative by using the above formula’s.

n
4nh f (t)
 
d n

1 X k n
(FDD) dt f (t) = lim = lim n (−1) f (t + (n − k)h)
h→0 hn h→0 h k
k=0
(3)
n
5nh f (t)
 
d n 1 X n
(−1)k


(BDD) dt f (t) = lim n
= lim n f (t − kh) (4)
h→0 h h→0 h k
k=0

Note that 4h f (t) = 5h f (t + h) and from the above formulas it follows that

4nh f (t) = 5nh f (t + nh) (5)

4
1.2.2 Cauchy formula for repeated integration
Rt
Integral of a function f (t) is denoted by a f (ν)dν. Integrating f (t) twice is
given by Z Z t ν1
f (ν2 )dν2 dν1
a a
If we use integration by parts we find
Z t Z ν1 Z t  Z ν1 
f (ν2 )dν2 dν1 = 1· f (ν2 )dν2 dν1
a a a a
h i Z ν1  ν1 =t Z th i
= − (t − ν1 ) f (ν2 )dν2 − − (t − ν1 ) f (ν1 )dν1
a ν1 =a a
Z t
= (t − ν1 )f (ν1 )dν1
a

In a similar manner we see for k ∈ Z, k ≥ 0

Z t  Z ν1 
k
(t − ν1 ) · f (ν2 )dν2 dν1
a a
 Z ν1 ν1 =t t
−(t − ν1 )k+1 −(t − ν1 )k+1
 Z
= f (ν2 )dν2 − f (ν1 )dν1
k+1 a ν1 =a a k+1
Z t
1
= (t − ν1 )k+1 f (ν1 )dν1 (6)
k+1 a
Would we integration f (t) n times and use the results above we would find
Z t Z ν1 Z νn−1 Z t
1
··· f (νn )dνn · · · dν1 = (t − ν)n−1 f (ν)dν (7)
a a a (n − 1)! a
| {z }
n

This formula is known as the Cauchy formula for repeated integration.

1.2.3 Gamma and Beta Function

The Gamma function Γ(z) is defined by
Z ∞
Γ(z) = tz−1 e−t dt
0

and extends the factorial by n! = Γ(n+1), for n ∈ Z, and Γ(z+1) = zΓ(z), z ∈ C.

The Beta function B(x, y) is defined by

Z 1
B(x, y) = tx−1 (1 − t)y−1 dt
0

where Re(x), Re(y) > 0.

The Gamma and Beta function share the following relation:

Γ(x)Γ(y)
B(x, y) =
Γ(x + y)

5
A different representation of the Gamma function can be found in Krantz 1999,
p.156
(n + 1)z n!
Γ(z) = lim
n→∞ z(z + 1) · · · (z + n)

There is no z ∈ C such that Γ(z) = 0. As a consequence, the reciprocal Gamma

1
function Γ(z) never tends to infinity on a compact subset of C. In addition,
1
Γ(k) = 0 for all k ∈ Z, k ≤ 0.

2 Differintegrals
In fractional calculus, a differintegal is an operator that connects differentia-
tion and integration. The operation of differentiation or integration using a
differintegral is called differintegration. It is important to note that there are
many ways to define a differintegral, as we shall see later on. In this section we
will discuss the Grünwald-Letnikov differintegral (represented by a Dtγ ) and the
Riemann-Liouville Differintegral (represented by a Dγt ).

In the notation, γ is the order of differintegration. For positive integers, γ

corresponds with regular differentiation. For negative integers, γ corresponds
with regular integration. The important part of defining differintegrals is mak-
ing sure that it also handles non-integer γ well.

a and t are called the terminals of the differintegral. We will get to them
later.

From time to time it might be unclear whether we are integrating of differenti-

ating, due to the fact that differintegrals can do both. To prevent this confusion
we define p, q ≥ 0 and use p and q to represent the order of differintegration
throughout the rest of the paper. p will be used when we are differentiating,
and q will be used when we are integrating.

2.1 Grünwald-Letnikov Differintegral ([2] p.43-48)

In this section we will discuss the Grünwald-Letnikov (GL) differintegral by con-
structing it from the bottom up. The GL differintegral is based on generalizing
the backward difference derivative (BDD). We shall see how this generalization
coincides with integration and extend it to arbitrary (non-integer) order.

2.1.1 Differentiation
Consider the real valued function f (t). Let f (t) be p times differentiable. Using
the BDD, the p-th derivative of f is given by
p  
d p

1 X k p
dt f (t) = lim (−1) f (t − kh)
h→0 hp k
k=0

6
Next we define the operator Q:
n  
1 X k p
Qph,n f (t) := p (−1) f (t − kh) (8)
h k
k=0
where p, n, k ∈ N and h ∈ R.

From this definition it follows that, for n ≥ p,

dp
lim Qph,n f (t) = f (p) (t) = dtp f
h→0

p


since k = 0 for k > p.

2.1.2 Integration
We will now show that for negative p this generalisation corresponds to regular
integration. Note that binomial coefficient can be expressed as
 
n n! n(n − 1)(n − 2) · · · (n − k + 1)
= =
k k!(n − k)! k!

We will define  
n n(n + 1) · · · (n + k − 1)
:= (9)
k k!
Plugging in negative values for n in the binomial coefficient we get
 
−n −n(−n − 1) · · · (−n − k + 1)
=
k k!
n(n + 1) · · · (n + k − 1)
= (−1)k
  k!
k n
= (−1)
k

Substituting this in (8) we find

n  
1 X −p
Qph,n f (t) = p f (t − kh)
h k
k=0

In particular, for p = −q,

n  
q
Q−q
X
q
h,n f (t) =h f (t − kh)
k
k=0

(−q)
Would we fix n and q and take the limit h → 0 we would find limh→0 fh,n = 0,
which is not very interesting. Instead we let n be defined by the relation nh =
t − a, such that n → ∞ as h → 0. As stated before, a and t are called the
terminals, and we will soon see that these form the region of integration. We
will denote all this as by
−q
a Dt f (t) := lim Q−q
h,n f (t)
h→0
nh=t−a

7
If we pick q = 1 we find
n  
−1
X 1
a Dt f (t) = lim Q−1
h,n f (t) = lim h f (t − kh)
h→0 h→0 k
nh=t−a nh=t−a k=0
n
X Z t−a
= lim h f (t − kh) = f (t − τ )dτ
h→0 0
nh=t−a k=0
Z t
= f (ν)dν
a

provided that f (t) is integrable. We shall prove by induction to q that in general

n   Z t
−q q
X q 1
a Dt f (t) = lim h f (t − kh) = (t − ν)q−1 f (ν)dν (10)
h→0 k (q − 1)! a
nh=t−a k=0

Recall the Cauchy formula for repeated integrals (7)

Z t Z t Z ν1 Z νq−1
1 q−1
(t − ν) f (ν)dν = ··· f (νq )dνq · · · dν1
(q − 1)! a
| a a {z a }
q

Proof Suppose we have for q = r that (10) holds. We will show that it follows
that (10) also holds for q = r + 1. We have
n  
−(r+1) (r+1)
X r+1
a Dt f (t) = lim h f (t − kh) (11)
h→0 k
nh=t−a k=0
Rt
Say f (−1) (t) = a
f (τ )dτ , then
1 h (−1) i
f (t) = lim f (t) − f (−1) (t − h)
h→0 h

Equivalently,
1 h (−1) i
f (t − kh0 ) = lim f (t − kh0 ) − f (−1) (t − kh0 − h)
h→0 h

If we take h = h0 and plug this into (11) we obtain

n  
−(r+1)
X r+1  
a Dt f (t) = lim hr f (−1) (t − kh) − f (−1) (t − (k + 1)h) (12)
h→0 k
nh=t−a k=0

From definition (9) it follows that

 
r+1 (r + 1)(r + 2) · · · (r + k)
=
k k!
(r + 1)(r + 2) · · · (r + k − 1)
= (r + k)
k!
r(r + 1) · · · (r + k − 1) (r + 1)(r + 2) · · · (r + k − 1)
= +
k! (k − 1)!
   
r r+1
= +
k k−1

8
  
r+1
Where we put = 0.
−1
Rt
Note that f (−1) (a) = 0 since f (−1) (t) = a f (τ )dτ . Using the above in (12),
along with nh = t − a, we get

n  
−(r+1)
X r+1  
a Dt f (t) = lim hr f (−1) (t − kh) − f (−1) (t − (k + 1)h)
h→0 k
nh=t−a k=0
n 
r + 1 (−1)
X
= lim hr f (t − kh)
h→0 k
nh=t−a k=0
n  
r + 1 (−1)
X 
− f (t − (k + 1)h)
k
k=0
n   n  
X r (−1) X r + 1 (−1)
= lim hr f (t − kh) + f (t − kh)
h→0 k k−1
nh=t−a k=0 k=0
n  
X r + 1 (−1) 
− f (t − (k + 1)h)
k
k=0
n   n  
r (−1) r + 1 (−1)
X X
= lim hr f (t − kh) + f (t − kh)
h→0 k k−1
nh=t−a k=0 k=1
n  
X r + 1 (−1) 
− f (t − (k + 1)h)
k
k=0
n   n−1
X 
X r r + 1 (−1)
= lim hr f (−1) (t − kh) + f (t − (k + 1)h)
h→0 k k
nh=t−a k=0 k=0
n  
X r + 1 (−1) 
− f (t − (k + 1)h)
k
k=0
n    
r (−1) r + 1 (−1)
X 
r
= lim h f (t − kh) − f (t − (n + 1)h)
h→0 k n
nh=t−a k=0
 
r + 1 (−1)
= a Dt−r f (−1) (t) − lim hr f (t − (n + 1)h)
h→0 n
nh=t−a
 
−r (−1) r r+1
= a Dt f (t) − lim h f (−1) (t − nh − h)
h→0 n
nh=t−a
 
−r (−1) r r+1
= a Dt f (t) − lim h f (−1) (t − (t − a) − h)
h→0 n
nh=t−a
 
r + 1 (−1)
= a Dt−r f (−1) (t) − lim hr f (a − h)
h→0 n
nh=t−a
 
r+1
= a Dt−r f (−1) (t) − f (−1) (a) lim hr
h→0 n
nh=t−a
−r (−1)
= a Dt f (t) −0

9
From this we deduce that
Z t
−(r+1) 1
a Dt f (t) = a Dt−r f (−1) (t) = (t − ν)r−1 f (−1) (ν)dν
(r − 1)! a
Z t Z ν
1
= (t − ν)r−1 f (τ )dτ dν
(r − 1)! a a
1 t
Z
= (t − ν)r f (ν)dν
r! a

where we use (6). This proves the claim. 

2.1.3 Arbitrary (Non-Integer) Order

So far we have defined
n  
γ 1 X k γ
a Dt f (t) := lim (−1) f (t − kh) (13)
h→0 hγ k
nh=t−a k=0

We have shown that for integer values of γ, a Dtγ f (t) corresponds to derivatives of
f when γ is positive and antiderivative of f when γ is negative. It also follows
from this definition that for γ = 0, a Dt0 f (t) = f (t). We now simply extend
this definition for non-integer order differintegrals by allowing non-integer γ in
the formula, substituting the Gamma function for the binomial product. The
existence of the limit is proved in [2] p.52-55.
Note that when γ is positive f (t) needs to be at least m times differentiable,
where m ≥ γ. When γ is negative, f (t) needs to be integrable.

2.1.4 Alternative Representation

There is an other way to express of the Grünwald-Letnikov derivative. From
the binomial coefficient we know that
 
p p! (k + p − k)! (p − 1)!k (p − 1)!(p − k)
= = = +
k k!(p − k)! k!(p − k)! k!(p − k)! k!(p − k)!
   
(p − 1)! (p − 1)! p−1 p−1
= + = + (14)
(k − 1)!(p − k)! k!(p − k − 1)! k−1 k

where we put p−1

−1 = 0.

10
Using this repeatedly in (8) we get
n  
(p) k p
X
p
h · fh,n (t) = (−1) f (t − kh)
k
k=0
(Equation (14))
n   n  
X p−1 X p−1
= (−1)k f (t − kh) + (−1)k f (t − kh)
k k−1
k=0 k=0
n   n  
X p−1 X p−1
= (−1)k f (t − kh) + (−1)k f (t − kh)
k k−1
k=0 k=1
n   n−1  
X p−1 X p−1
= (−1)k f (t − kh) + (−1)k+1 f (t − (k + 1)h)
k k
k=0 k=0
n−1  h  
k p−1 n p−1
X i
= (−1) f (t − kh) − f (t − (k + 1)h) + (−1) f (t − nh)
k n
k=0
n−1    
k p−1 n p−1
X
= (−1) 5h f (t − kh) + (−1) f (a)
k n
k=0
(Repeat once more)
n−2  
X p−2 h i
= (−1)k 5h f (t − kh) − 5h f (t − (k + 1)h)
k
k=0
   
p−2 p−1
+ (−1)n−1 5h f (t − (n − 1)h) + (−1)n f (a)
n−1 n
n−2  
X p−2
= (−1)k 52h f (t − kh)
k
k=0
   
n−1 p − 2 n p−1
+ (−1) 5h f (a + h) + (−1) f (a)
n−1 n
(After m times, m ≤ p ≤ n)
n−m   m−1  
k p−m n−l p − 1 − l
X X
m
= (−1) 5h f (t − kh) + (−1) 5lh f (a + lh)
k n−l
k=0 l=0
(Using identity (5))
n−m   m−1  
X p−m X
n−l p − 1 − l
= (−1)k 5m
h f (t − kh) + (−1) 4lh f (a)
k n−l
k=0 l=0

Thus we see that

p (p)
a Dt f (t) = lim fh,n (t) =
h→0
nh=t−a
n−m  
1 X k p−m
= lim (−1) 5m
h f (t − kh)
h→0 hp k
nh=t−a k=0
m−1  
1 X p−1−l
+ lim (−1)n−l 4lh f (a)
h→0 hp n−l
nh=t−a l=0

11
 (n+1)z n!
Using equations (10), (4), (3) and the identity Γ(z) = limn→∞ z(z+1)···(z+n) , we
find that
t m−1
f (l) (a)(t − a)−p+l
Z
p 1 X
a Dt f (t) = (t − ν)m−p−1 f (m) (ν)dν +
Γ(−p + m) a Γ(−p + l + 1)
l=0
(15)

We omit the full proof of this statement. All details of this construction can be
found in [2] p.52-55.
For this representation we see that f (t) needs to be at least m times differen-
tiable.

The Grünwald-Letnikov differintegral is a nice definition for fractional inte-

grals and derivatives, but it is not a quick tool to determine the representation
of a derivative or integral of arbitrary order. However, it is very suitable for
numerical approximations ([2] p.199 and on).

2.2 Riemann-Liouville Differintegral

The Riemann-Liouville (RL) differintegral is very similar to the Grünwald-
Letnikov differintegral. It is based on extending the Cauchy formula for repeated
integration (see (7)) for non-integer values, in order to define fractional integra-
tion. Fractional differentiation is defined by first applying fractional integration
and afterwards performing regular differentiation, in a unique way.

To illustrate this concept, we picture a “time line” of γ, where γ is the or-

der of differintegration. Positive values of γ correspond with differentiation.
Negative values of γ correspond with integration.

γ −3 −2 −1 0 1 2 3

f (γ) f (-3) f (-2) f (-1) f (0) f (1) f (2) f (3)

Lets interpret the rules above. First we take a move to the left, then we move
to the right. We can take arbitrary size steps to the left on our time line, but
only integer steps to the right. Thus, to calculate a fractional derivative of f ,
we first have to move a certain amount to the left and subsequently move with
integer steps to the right.
For example, would we like to know the result for γ = 3/2, we first have to take
a half step to the left and subsequently two steps to the right.

γ −3 −2 −1 0 1 2 3

f (γ) f (-3) f (-2) f (-1) f (0) f (1) f (2) f (3)

Obviously, this can be done in multiple ways, like taking one and a half steps to
the left and three steps to the right. In order to make the fractional derivative
unique we require the step to the left to be between 0 and -1, thus of size less
than 1.

12
2.2.1 Fractional Integral
The Riemann-Liouville definition of the fractional integral is given by
Z t
−q 1
D
a t f (t) = (t − ν)q−1 f (ν)dν
Γ(q) a
where q ∈ R+ and where we put a D0t f (t) = f (t). For details on why we can
put a D0t f (t) = f (t), see [2] p.65.

It is merely the non-integer extension of Cauchy formula for repeated integra-

tion (see (7)), substituting the Gamma function for the factorial. Note that we
need f (t) to be integrable.

As a warning, in Miller-Ross ([1]) this definition is referred to as the Rie-

mann version of the fractional integral, whereas they say the Riemann-Liouville
version is where a = 0.

2.2.2 Fractional Derivative

For the fractional derivative we decompose the fractional order p by p =: m − q,
where m ∈ Z+ and q ∈ (0, 1). If p is an integer we will use the regular derivative.
To calculate the fractional derivative we first preform a fractional integration of
order q, following regular differentiation of order m. Thus, we define that the
Riemann-Liouville definition of the fractional derivative is given by
d m Z t


p d m −q dt
(t − ν)q−1 f (ν)dν


a Dt f (t) = dt a Dt f (t) =
Γ(q) a
In contrast with the GL differintegral, we only require f (t) to be integrable and
we don’t need that f (t) is m times differentiable.

2.2.3 The Fractional Integral and the Leibniz Integral Rule

Theorem 2.1 (Leibniz Integral Rule) Consider the function f (t, ν) and the
R b(t)
integral a(t) f (t, ν)dν. Let f (t, ν) be a continuous function that has a contin-
uous derivative with respect to t within the region of integration. Suppose that
a(t) and b(t) are also continuous and have continuous derivatives inside this
region. Then, if t is inside this region
 Z b(t)  Z b(t)
d 0 0 d
f (t, ν)dν = f (x, b(x))b (x) − f (x, a(x))a (x) + f (t, ν)dν
dt a(t) a(t) dt

Applying the theorem above to our definition of the fractional integral, we find
that for q > 1
Z t
(t − t)q−1
 
d 1
(t − ν)q−1 f (ν)dν = ·1−0
dt Γ(q) a Γ(q)
Z t 
1 d
+ (t − ν)q−1 f (ν)dν
Γ(q) a dt
Z t
1
= (t − ν)q−2 f (ν)dν
Γ(q − 1) a

13
In general, for q > 1 and m ≥ 1,
 m  Z t  Z t
d 1 q−1 1
(t − ν) f (ν)dν = (t − ν)q−1−m f (ν)dν
dt Γ(q) a Γ(q − m) a

Or in notation
d m −q
= a Dm−q


dt a Dt f (t) t f (t), q > 1 and m ≥ 1 (16)

In addition, this holds for non-integer values of q provided that t − ν ≥ 0,

implying t ≥ a.

2.2.4 Interchanging Derivation and Fractional Integration

d d
Note that dt (t − ν)α = − dν (t − ν)α . We will use this a lot in the rest of the
paper when we are using integration by parts.

Taking the m-th derivative of a fractional integral a D−q t f (t) we get

d m Z t


d m −q dt
(t − ν)q−1 f (ν)dν


dt a Dt f (t) =
Γ(q) a
d m−1  Z t h

i 
dt d q−1 q−1
=
Γ(q) dt (t − ν) f (ν)dν + lim (t − ν) f (ν)
ν→t
a
d

m−1  Z th i 
= dt − dνd
(t − ν)q−1 f (ν)dν + lim (t − ν)q−1 f (ν)
Γ(q) a ν→t

(integration by parts)
d m−1  Z t h

i
= dt
(t − ν)q−1 f (1) (ν)dν
Γ(q) a
ν=t 
−(t − ν)q−1 f (ν) + lim (t − ν)q−1 f (ν)
ν→t
ν=a
d m−1  Z t


= dt
(t − ν)q−1 f (1) (ν)dν + (t − a)q−1 f (a)
Γ(q) a
d m−1 Z t


(t − a)q−m f (a)
= dt
(t − ν)q−1 f (1) (ν)dν +
Γ(q) a Γ(q − m + 1)
(repeat once more)
d m−2 Z t


(t − a)q−m+1 f (1) (a) (t − a)q−m f (a)
= dt
(t − ν)q−1 f (2) (ν)dν + +
Γ(q) a Γ(q − m + 2) Γ(q − m + 1)
(after m times)
Z t m−1
X (t − a)q−m+l f (l) (a)
1 q−1 (m)
= (t − ν) f (ν)dν + (17)
Γ(q) a Γ(q − m + l + 1)
l=0
m−1
(t − a)q−m+l f (l) (a)
=a D−q
X
(m)
t f (t) + (18)
Γ(q − m + l + 1)
l=0

From this we deduce the following theorem:

14
Theorem 2.2 Let f (t) be an m times differentiable function. Then
d m −q −q (m)


dt a t f (t) = a Dt f
D (t) if and only if f (l) (a) = 0 for 0 ≤ l ≤ m−1.

If we recall our time line example, the above statement implies that it doesn’t
matter whether we step to the left or to right first if and only if f (l) (a) = 0 for
0 ≤ l ≤ m − 1, provided that f (t) is m times differentiable.

2.2.5 Relation to the Grünwald-Letnikov Differintegral

We will show that the Riemann-Liouville differintegral equals the Grünwald
-Letnikov differintegral. We already know that a D0t f (t) = f (t) = a Dt0 f (t).
From (10) it follows that a Dt−q f (t) = a Dt−q f (t) for q ∈ R+ , provided that f (t)
is integrable.
Now for p ∈ R+ , we decompose p by p =: m − q, m ∈ Z+ and q ∈ (0, 1). Let
f (t) be m times differentiable. Using equation (17) we see that
p d m −q


a Dt f (t) = dt a Dt f (t) =
m−1
X (t − a)q−m+l f (l) (a) Z t
1
= + (t − ν)q−1 f (m) (ν)dν
Γ(q − m + l + 1) Γ(q) a
l=0
m−1
X (t − a)−p+l f (k) (a) Z t
1
= + (t − ν)m−p−1 f (m) (ν)dν
Γ(−p + l + 1) Γ(−p + m) a
l=0

Comparing this result with equation (15) we find

m−1 t
(t − a)−p+l f (k) (a)
Z
p
X 1
a Dt f (t) = + (t − ν)m−p−1 f (m) (ν)dν
Γ(−p + l + 1) Γ(−p + m) a
l=0
= a Dtp f (t)

We can conclude that a Dγt f (t) = a Dtγ f (t) for all γ ∈ R, provided that f (t) is
integrable and m times differentiable.

3 Backwards Compatibility
It should be stressed once more that there are many different approaches to
extending integrals and derivatives to non-integer order. In the literature the
Riemann-Liouville definition is the most used version of the differintegral oper-
ator and the most commonly accepted fractional integral and fractional deriva-
tive. Therefore we will be using the RL differintegral throughout the end of the
paper.

In this section we will check whether the RL differintegral follows the regu-
lar rules of the differentiation and integration. We will start by checking under
which conditions we can interchange fractional integration and fractional differ-
entiation. We will also discuss linearity and the product rule.

15
3.1 Integral and Derivative Compositions
k l l k (k+l)
d d d d d
From calculus we know that dt k dtl f (t) = dtl dtk f (t) = dt(k+l) f (t), provided

that f (t) is k + l times differentiable. In this section we will check under which
conditions compositions of fractional derivatives and integrals still commute and
are still interchangeable.

3.1.1 Integral of an Integral

Let f (t) be integrable. We will look at the composition of two fractional inte-
grals.
Z t Z ν
−q1 −q2 1 q1 −1
a Dt [a Dt f (t)] = (t − ν) (ν − µ)q2 −1 f (µ)dµdν
Γ(q1 )Γ(q2 ) a a
Z ν Z t
1 q2 −1
= (F ubini) (ν − µ) (t − ν)q1 −1 f (µ)dνdµ
Γ(q1 )Γ(q2 ) a a

Change of coordinates: ν = µ + γ(t − µ). The boundaries of integration become

( ( ( (
µ =ν µ =t µ =a µ =a
=⇒ =⇒
ν =t γ =1 ν =t γ =1
( ( ( (
µ =ν µ =a µ =a µ =a
=⇒ =⇒
ν =a γ =0 ν =a γ =0

Thus [a, ν] × [a, t] =⇒ [a, t] × [0, 1]

Z tZ 1
−q1 1
a Dt [a D−q
t
2
f (t)] = γ q2 −1 (t − µ)q2 −1
Γ(q1 )Γ(q2 ) a 0
·(1 − γ)q1 −1 (t − µ)q1 −1 f (µ) dγ (t − µ) dµ
Z tZ 1 
1 q2 −1 q1 −1
= γ (1 − γ) dγ (t − µ)q1 +q2 −1 f (µ) dµ
Γ(q1 )Γ(q2 ) a 0
Z tZ 1 
1
= γ q2 −1 (1 − γ)q1 −1 dγ (t − µ)q1 +q2 −1 f (µ) dµ
Γ(q1 )Γ(q2 ) a 0
Z t
1
= B(q1 , q2 )(t − µ)q1 +q2 −1 f (µ) dµ
Γ(q1 )Γ(q2 ) a
Z t
B(q1 , q2 )
= (t − µ)q1 +q2 −1 f (µ) dµ
Γ(q1 )Γ(q2 ) a
Z t
1
= (t − µ)q1 +q2 −1 f (µ) dµ
Γ(q1 + q2 ) a
−(q1 +q2 )
= a Dt f (t)

We see that
−q1 −(q1 +q2 )
a Dt [a D−q
t
2
f (t)] = a D−q
t
2
[a D−q
t
1
f (t)] = a Dt f (t) (19)

So as long as f (t) is integrable, negative orders can be added.

16
3.1.2 Derivative of an Integral
We will be looking at the fractional derivative of a fractional integral.
Let p1 =: m1 − q1 , where m1 ∈ Z+ and q1 ∈ (0, 1), and let q2 ∈ Z+ .
p1 −q2 d m1 −q1 −q2


a D t [a D t f (t)] = dt a Dt a Dt f (t)

Using (19) to combine the integrals and (16) to reduce the order of the integral
such that it is inside (0, 1), we see that
d m1 −q1 −q2 d m1 −(q1 +q2 )



dt a Dt a Dt f (t) = dt a Dt f (t)
d

m 1 d

q3 −(q 1 +q 2 ) −q 3 d

m1 +q 3 −(q1 +q2 ) −q3
= dt dt a Dt f (t) = dt a Dt f (t)
d m3 −q3 m3 −q3 p1 −q2


= dt a Dt f (t) = a Dt f (t) = a Dt f (t) (20)

where m3 := m1 + q3 − (q1 + q2 ) and q3 ∈ (0, 1) such that m3 ∈ Z+ . So

−q2
p1
a Dt [a Dt f (t)] = a Dpt 1 −q2 f (t)

3.1.3 Integral of a Derivative

We will look at the fractional integral of a fractional derivative.
Let p2 =: m2 − q2 , where m2 ∈ Z+ and q2 ∈ (0, 1), and let q1 ∈ Z+ . First we
observe that
−q1 d m2
[a Dpt 2 f (t)] = a Dt−q1 dt −q2


a Dt a Dt f (t)

If we apply theorem 2.2 to this formula, and subsequently (18), we see that
−q1 d
m2 −q2 d m2

−q1 −q2
a Dt dt a Dt f (t) = dt a Dt a Dt f (t)
⇐⇒
d l −q2


dt a Dt f (t) =0 at t = a, 0 ≤ l ≤ m2 − 1
⇐⇒
l−1 q−l+k (k)
−q (l)
X (t − a) f (a)
a Dt f (t) + =0 at t = a, 0 ≤ l ≤ m2 − 2
Γ(q − l + k + 1)
k=0

Now a D−q (l)

t f (t) = 0 at t = a, since it is an integral over zero length. This
Pl−1 (t−a)q−l+k f (k) (a)
implies k=0 Γ(q−l+k+1) = 0, which implies that f (l) (a) = 0 for 0 ≤ l ≤
m2 − 2.

We end up with the following:

−q1 d
m2 −q2 d m2

−q1 −q2
a Dt dt a Dt f (t) = dt a Dt a Dt f (t)
if and only if
f (l) (a) = 0 for 0 ≤ l ≤ m2 − 2

Now we can use (20) on the right hand side of the equation such that we find
−q1
a Dt [a Dpt 2 f (t)] = a Dtp2 −q1 f (t) (21)

provided that f (l) (a) = 0 for 0 ≤ l ≤ m2 − 2.

17
3.1.4 Derivative of a Derivative
Let p1 =: m1 − q1 and p2 =: m2 − q2 , where m1 , m2 ∈ Z+ and q1 , q2 ∈ (0, 1).
Using (21) and (16) we see that
p1 p2 d m1 −q1
[a Dpt 2 f (t)]


a Dt [a Dt f (t)] = dt a Dt
d m1 p2 −q1


= dt a Dt f (t)
= a Dpt 1 +p2 f (t)

under the assumption that f (l) (a) = 0 for 0 ≤ l ≤ m2 − 2.

Similarly, would we consider interchanging the two derivatives we end up with

p2 p1
a Dt [a Dt f (t)] = a Dpt 1 +p2 f (t)

provided that f (l) (a) = 0 for 0 ≤ l ≤ m1 − 2. Let r be defined by

r := max(m1 − 2, m2 − 2). We find
p1 p2
a Dt [a Dt f (t)] = a Dpt 2 [a Dpt 1 f (t)] = a Dpt 1 +p2 f (t)
⇐⇒
(l)
f (a) = 0 for 0 ≤ l ≤ r

We need f (t) to be at least r times differentiable in this case.

3.2 Linearity
From the linearity of the integral it follows that a D−q
t is linear.
Z t
−q 1
a Dt (cf (t) + dg(t)) = (t − ν)q−1 (cf (ν) + dg(ν))dν
Γ(q) a
Z t Z t
1 q−1 1
=c (t − ν) f (ν)dν + d (t − ν)q−1 g(ν)dν
Γ(q) a Γ(q) a
= c a D−q −q
t f (t) + d a Dt g(t)

In a similar fashion
p d m −q


a Dt (cf (t) + dg(t)) = dt a Dt (cf (t) + dg(t))
h i
d m
c a D−q −q


= dt t f (t) + d a Dt g(t)
d m −q d m −q



= c dt a Dt f (t) + d dt a Dt g(t)
= c a Dpt f (t) + d a Dpt g(t)

3.3 Product Rule and Leibniz Rule

d d d
We know the product rule as dt (f (t) · g(t)) = f (t) · dt g(t) + g(t) · dt f (t).
This gives rise to the Leibniz rule:
n  h
d n

X n d k

ih
d n−k

i
dt (f (t) · g(t)) = dt f (t) dt g(t)
k
k=0

18
 n


Since k = 0 for k > n we can extend the summation for any r ≥ n:
r  h
d n

X n d k

ih

d n−k
i
dt (f (t) · g(t)) = dt f (t) dt g(t)
k
k=0

The Leibniz rule for fractional derivatives will be defined by

r  h
p
X p d k

ih
p−k
i
a Dt (f (t) · g(t)) = dt f (t) a Dt g(t)
k
k=0
d m −q
Since a Dpt f (t) = dt


a Dt f (t) we may also write
r  h
p
X p d k

ih

d m−k −q
i
D
a t (f (t) · g(t)) = dt f (t) dt a Dt g(t)
k
k=0

The reasoning behind why we may define the Leibniz rule for fractional deriva-
tives this way are given in [2] p.91-97.
The chain rule can also be expanded for non-integer orders, but we omit it here
due to its complexity. Those who are interested can find it in [2] p.97-98.

4 Examples of Differintegration of Functions

This section will feature fractional derivatives of some basic function to give
an indication how they are conceived. We will be using the RL differintegral
throughout the end of the paper.

4.1 The Power function

Let f (t) = tr , and a = 0. Let α ≤ r. Then the differintegral of f is given by
d m Z t


α
a Dt f (t) = 0 Dα r
tt =
dt
(t − ν)q−1 ν r dν
Γ(q) 0
Using the coordinate transformation ν = γt we find
d m Z t d m Z 1



q−1 r
dt
(t − ν) ν dν = dt
(t − γt)q−1 (γt)r t dγ
Γ(q) 0 Γ(q) 0
d m Z 1


= dt
tq−1 (1 − γ)q−1 γ r tr+1 dγ
Γ(q) 0
d m 1

Z
q+r
= dt
t (1 − γ)q−1 γ r dγ
Γ(q) 0
Z 1
Γ(q + r + 1)
= tq−m+r (1 − γ)q−1 γ r dγ
Γ(q)Γ(q − m + r + 1) 0
Γ(q + r + 1)
= B(q, r + 1)tq−m+r
Γ(q)Γ(q − m + r + 1)
Γ(q + r + 1) Γ(q)Γ(r + 1) q−m+r
= t
Γ(q)Γ(q − m + r + 1) Γ(q + r + 1)
Γ(r + 1) Γ(r + 1) r−α
= tq−m+r = t
Γ(q − m + r + 1) Γ(r − α + 1)
provided that r > −1

19
 Plots for 0 Dα
tt
3/2
, with t ∈ [0, 5] and α ∈ [−3, 0]

4.2 The Constant Function

To obtain the differintegral of a constant we use the expression we found above
and take the limit of r to 0. We find
α Γ(r + 1) r−α t−α
0 Dt 1 = lim 0 Dα r
t t = lim t =
r→0 r→0 Γ(1 + r − α) Γ(1 − α)
And due to the linearity of the RL differintegral (see section 3.2)

α Ct−α
0 Dt C=
Γ(1 − α)
1
An important property of the reciprocal gamma function Γ(z) is that it equals
zero if z is an integer less or equal to zero. From this it follows that the derivative
of a constant is zero whenever the order α is an integer greater than zero,
and non-zero in between these integer values. Though the first outcome is
exactly what we expected due to the construction of our differintegral, the
second outcome is rather surprising.

α
Plots for −∞ Dt 1, with t ∈ [0, 5] and α ∈ [0, 5]

20
4.3 The Exponential Function
Let f (t) = ert with r ∈ C, and a = −∞. An RL differintegral where a = −∞
is known as a Weyl differintegral.
d m Z t


α
a Dt f (t) = −∞ Dα rt
t e =
dt
(t − ν)q−1 erν dν
Γ(q) 0

Using the coordinate transformation µ = r(t − ν)

d m Z t d m Z t



dt q−1 rν 1−q dt
q−1 rν
(t − ν) e dν = r r(t − ν) e dν
Γ(q) 0 Γ(q) 0
d m Z ∞ d m rt Z ∞



e
= r1−q dt µq−1 ert−µ r−1 dµ = r−q dt µq−1 e−µ dµ
Γ(q) 0 Γ(q) 0
d m rt


−q dt e d m rt
Γ(q) = r−q dt e = r−q rm ert = rα ert


=r
Γ(q)
In short,
α
−∞ Dt ert = rα ert (22)
α
In particular, −∞ Dt et = et for all α.

α 2t
Plot for −∞ Dt e , with t ∈ [−3, 5] and α ∈ [0, 4]

4.4 Trigonometric Functions

To calculate differintegals for trigonometric functions we express them using the
exponential function with the following formulas
eit − e−it eit + e−it
sin t = cos t =
2i 2
In order to use equation (22) we put a = −∞. Recalling the linearity of the RL
differintegral (section 3.2), we find
h it −it i
α e −e 1h i
α α it α −it
−∞ Dt sin t = −∞ Dt = −∞ Dt e − −∞ Dt e
2i 2i
1 h α it α −it
i 1 h i π α it π
i
= i e − (−i) e = e 2 e − e−i 2 α e−it
2i 2i
1 h i(t+ π α) π
i
= e 2 − e−i(t+ 2 α) = sin(t + π2 α)
2i
α
Likewise, for the cosine we find −∞ Dt cos t = cos(t + π2 α).

21
 α α
Plots for −∞ Dt sin(t) and −∞ Dt cos(t), with t ∈ [0, 4π] and α ∈ [0, 4]

4.5 Hyperbolic Functions

Analogue to the trigonometric functions we can find an expression for differin-
tegrals of hyperbolic functions.

et − e−t et + e−t
sinh t = cosh t =
2 2
For the hyperbolic sine function
 et − e−t 1 
α α −t
−∞ Dt sinh t = −∞ Dα
t = α t
−∞ Dt e − −∞ Dt e
2 2
1 t α −t
 1 
= e − (−1) e = et − eiπα e−t
2 2
1 i π α  −i π α t π
−t
 1 π  π π

= e 2 e 2 e −e 2 e i α
= e 2 e 2 α − e−(t−i 2 α)
i α t−i
2 2
π
= ei 2 α sinh(t − i π2 α) = iα sinh(t + 2i
π
α)

Likewise, for the hyperbolic cosine we find

α π
−∞ Dt cosh t = ei 2 α cosh(t − i π2 α) = iα cosh(t + π
2i α)

d d
In addition, since dt sinh t = cosh t and dt cosh t = sinh t, the differintegrals
above give rise to the following identities:

sinh t = (−1)k sinh(t ± iπk)

cosh t = (−1)k cosh(t ± iπk)

sinh t = (−1)k i cosh(t ± iπk + π

2i )
cosh t = (−1)k i sinh(t ± iπk + π
2i )

for all k ∈ Z.

22
 α
Plots for the real part of −∞ Dt sinh(t) with t ∈ [−10, 10] and α ∈ [0, 8]

α
Plots for the imaginary part of −∞ Dt sinh(t) with t ∈ [−10, 10] and α ∈ [0, 8]

α
Plots for the real part of −∞ Dt cosh(t) with t ∈ [−10, 10] and α ∈ [0, 8]

23
 α
Plots for the imaginary part of −∞ Dt cosh(t) with t ∈ [−10, 10] and α ∈ [0, 8]

5 Discussion
We have seen different methods to extend regular integration and differentiation
to non-integer orders. Using the forward difference derivative and the Cauchy
formula for repeated integration we have defined the Grünwald-Letnikov and
Riemann-Liouville differintegrals. Although the differintegrals are rather differ-
ent, they give the same result, provided some conditions are met.

The RL differintegral is linear in the argument and the orders add up nicely
under the right conditions. Furthermore, we have calculated differintegrals of
some basic functions using the RL differintegral.

5.1 Other types of differintegrals

There are many other ways to represent integrals and differentials of functions.
This results in various definitions of differintegrals.

5.1.1 Caputo Differintegral

The Caputo differintegral is similar to the RL differintegral. The fractional
integral is also based on the Cauchy formula for repeated integration. In contrast
with the RL differintegral, the fractional derivative is defined by first taking a
regular derivative and then taking a fractional integral. As a result, all the
fractional derivatives of a constant are zero.

5.1.2 Differintegrals based on Laplace Transform

Rt
A different approach involves the Laplace transform. If F (t) =
0 f (ν)dν, then
the Laplace transform of F (t) is given by L{F (t)} = 1s L{f (t)} (s). From this
one may derive that
 Z t Z ν1 Z νn−1 
1

L ··· f (νn )dνn · · · dν1 = n L{f (t)} (s)
s
| a a {z a }
n

24
Or, taking the inverse Laplace transform of both sides
Z t Z ν1 Z νn−1
1
f (νn )dνn · · · dν1 = L−1 n L{f (t)} (s)


···
s
| a a {z a }
n

The fractional integral may be obtained by putting in non-integer values for n

on the right hand side. If f (0) = 0 and all derivatives of f (t) upto order n − 1
are zero at t = 0, the Laplace transform of the derivative of f (t) is
dn n


L{ dt n f (t)} = s L{f (t)} (s)

which is also easy to extend to non-integer order.

Under certain conditions this approach coincides with the RL differintegral,

as seen in [2] p.104-71. On the other hand, its easy to describe fractional
integrals using the Laplace transformation, as seen in [1] p.69-71.

5.1.3 Hadamard Fractional Integral

The Hadamard differintegral is given by
Z t q−1
−q 1 t dν
a Dt f (t) = log f (ν)
Γ(q) a ν ν

It is studied very little compared to the other differintegrals.

5.2 Subjects of Personal Interest

There are still quite a lot of things I would like to learn about fractional cal-
culus. For example, let a Df (t, α) = a Dα t f (t). What would it mean if we
let α be a function of t? Can we say that −∞ Dcos (t, − π2 t) = 1, since we
have −∞ Dα π
t cos t = cos(t + 2 α)? And, is there a function h(t) that satisfies
a Dh (t, α) = a Dh (α, t)?

With the concept of fractional order derivatives came also the notion of frac-
tional differential equations: differential equations using differintegal operators.
I was wondering whether there are cases in which it would be viable to split an
ordinary differential equation into derivatives of fractional order.

25
References
[1] Miller, K.S. and Ross, B.
An Introduction to the Fractional Calculus and Fractional Differential Equa-
tions
Isbn 9780471588849
Wiley, 1993
[2] Podlubny, I.
Fractional Differential Equations: An Introduction to Fractional Derivatives,
Fractional Differential Equations, to Methods of Their Solution and Some
of Their Applications
Isbn 9780080531984
Elsevier Science, 1998
[3] Das, S.
Functional Fractional Calculus
Isbn 9783642205453
Springer-Verlag Berlin Heidelberg, 2011

26