Вы находитесь на странице: 1из 10

International J.Math. Combin. Vol.

2 (2009), 01-10
Tangent Space and Derivative Mapping
on Time Scale
Emin

OZYILMAZ
(Department of Mathematics,University of Ege, 35100,

Izmir, Turkey)
E-mail: emin.ozyilmaz@ege.edu.tr
Abstract: A pseudo-Euclidean space, or Smarandache space is a pair (R
n
, |
O
). In this
paper, considering the time scale concept on Smarandache space with |
O
(u) = 0 for u E,
i.e., the Euclidean space, we introduce the tangent vector and some properties according to
directional derivative, the delta dierentiable vector elds on regular curve parameterized by
time scales and the Jacobian matrix of -completely delta dierentiable two variables function.
Key Words: Pseudo-Euclidean space, Smarandache space, time scale, regular curve,
derivative mapping.
AMS(2000): 53B30, 51B20.
1. Introduction
A pseudo-Euclidean space, or Smarandache space is a pair (R
n
, |
O
), where |
O
: R
n
O is a
continuous function, i.e., a straight line with an orientation

O will has an orientation

O+|
O
(u)
after it passing through a point u E. It is obvious that (E, |
O
) = E the Euclidean space
if and only if |
O
(u) = 0 for u E, on which calculus of time scales was introduced by
Aulbach and Hilger [1,2]. This theory has proved to be useful in the mathematical modeling of
several important dynamic processes [3,4,5]. We know that the directional derivative concept
is based on for some geometric and physical investigations. It is used at the motion according
to direction of particle at the physics [6]. Then, Bohner and Guseinov has been published
a paper about the partial dierentiation on time scale [7]. Here, authors introduced partial
delta and nabla derivative and the chain rule for multivariable functions on time scale and
also the concept of the directional derivative. Then, the directional derivative according to the
vector eld has dened [8]. The general idea in this paper is to investigate some properties of
directional derivative. Then, using the directional derivative, we dene tangent vector space
and delta derivative on vector elds. Finally, we write Jacobian matrix and the -derivative
mapping of the -completely delta dierentiable two variables functions. So our intention is to
use several new concepts, which are dened in dierential geometry [7].
1
Received Feb.25, 2009. Accepted April 10, 2009.
2 Emin

OZYILMAZ
2. Partial dierentiation on time scale
Let n N be xed. Further, for each i {1, 2, , n} let T
i
denote a time scale, that is, T
i
is
a nonempty closed subset of the real number R. Let us set

n
= T
1
xT
2
x xT
n
= {t = (t
1
, t
2
, , t
n
) for t
i
T
i
, 1 i n}
We call
n
an n-dimensional time scale. The set
n
is a complete metric space with the metric
d dened by
d(t, s) =

_
n

i=1
|t
i
s
i
|
2
, for t, s
n
.
Let
i
and
i
denote , respectively, the forward and backward jump operators in T
i
. Re-
member that for u T
i
the forward jump operator
i
: T
i
T
i
is dened by

i
() = inf{ T
i
: > }
and the backward jump operator
i
: T
i
T
i
by

i
() = inf{ T
i
: < }.
In this denition we put
i
(maxT
i
) = maxT
i
if T
i
has a nite maximum, and
i
(minT
i
) = minT
i
if T
i
has a nite minimum. If
i
() > , then we say that is right-scattered ( in T
i
), while
any with is left-scattered ( in T
i
). Also, if u < maxT
i
and
i
() = , then is called
right-dense ( in T
i
), and if > minT
i
and
i
() = , then is called left- dense ( in T
i
). If T
i
has a left-scattered minimum m, then we dene T
k
i
= T
i
{m}, otherwise T
k
i
= T
i
. If T
i
has
a right-scattered maximum M, then we dene (T
i
)
k
= T
i
{M}, otherwise (T
i
)
k
= T
i
.
Let f :
n
R be a function. The partial delta derivative of f with respect to t
i
T
k
i
is
dened as the limit
lim
s
i
t
i
si=i(ti)
f(t
1
, , t
i1
,
i
(t
i
), t
i+1
, , t
n
) f(t
1
, , t
i1
, s
i
, t
i+1
, , t
n
)

i
(t
i
) s
i
=
f(t)

i
t
i
.
Denition 2.1 We say a function f :
n
R is completely delta dierentiable at a point
t
0
= (t
0
1
, t
0
2
, , t
0
n
) T
k
1
xT
k
2
x xT
k
n
if there exist numbers A
1
, A
2
, , A
n
independent of
t = (t
1
, t
2
, , t
n
)
n
(but in general, dependent on t
0
) such that for all t U

(t
0
),
f(t
0
1
, t
0
2
, , t
0
n
) f(t
1
, t
2
, , t
n
) =
n

i=1
A
i
(t
0
i
t
i
) +
n

i=1

i
(t
0
i
t
i
)
and, for each j {1, 2, , n} and all t U

(t
0
),
Tangent Space and Derivative Mapping on Time Scale 3
f(t
0
1
, , t
0
j1
,
j
(t
0
j
), t
0
j+1
, , t
0
n
) f(t
1
, , t
j1
, t
j
, t
j+1
, , t
n
)
= A
j
(
j
(t
0
j
) t
j
) +
n

i=1
i=j
A
i
(t
0
i
t
i
) +
jj
(
j
(t
0
j
) t
j
) +
n

i=1
i=j

ij
(t
0
i
t
i
), (2 1)
where is a suciently small positive number, U

(t
0
) is the -neighborhood of t
0
in
n
,
i
=

i
(t
0
, t) and
ij
=
ij
(t
0
, t) are dened on U

(t
0
) such that they are equal to zero at t = t
0
and
such that
lim
tt
0

i
(t
0
, t) = 0 and lim
tt
0

ij
(t
0
, t) = 0 for all i, j {1, 2, , n}.
Denition 2.2 We say that a function f : T
1
xT R is
1
-completely delta dierentiable
at a point (t
0
, s
0
) T
1
xT if it is completely delta dierentiable at that point in the sense of
conditions (2.5)-(2.7)(see in [7]) and moreover, along with the numbers A
1
and A
2
presented in
(2.5)-(2.7) (see in [7]) there exists also a number B independent of (t, s) T
1
xT (but, generally
dependent on (t
0
, s
0
)) such that
f(
1
(t
0
),
2
(s
0
)) f(t, s)
= A
1
(
1
(t
0
) t) +B(
2
(s
0
) s) +
1
(
1
(t
0
) t) +
2
(
2
(s
0
) s) (2 2)
for all (t, s) V
1
(t
0
, s
0
), a neighborhood of the point (t
0
, s
0
) containing the point (
1
(t
0
), s
0
),
and the functions
1
=
1
(t
0
, s
0
, t, s) and
2
=
2
(t
0
, s
0
, t, s) are equal to zero for (t, s) = (t
0
, s
0
)
and
lim
(t,s)(t
0
,s
0
)

1
=
1
(t
0
, s
0
, t, s) and lim
ss
0

2
(t
0
, s
0
, s) = 0. (2 3)
Note that in (2 1) the function
2
depends only on the variable s. Setting s =
1
(s
0
) in
(2 2) yields
B =
f(
1
(t
0
), s
0
)

2
s
.
Furthermore, let two functions
: T R and : T R
be given and let us set
(T) = T
1
and (T) = T
2
.
We will assume that T
1
and T
2
are time scales. Denote by
1
,
1
and
2
,
2
the forward
jump operators and delta operators for T
1
and T
2
, respectively. Take a point
0
T
k
and put
t
0
= (
0
) and s
0
= (
0
).
4 Emin

OZYILMAZ
We will also assume that
((
0
)) =
1
((
0
)) and ((
0
)) =
2
((
0
)).
Under the above assumptions let a function f : T
1
xT
2
R be given.
Theorem 2.1 Let the function f be
1
-completely delta dierentiable at the point (t
0
, s
0
). If
the functions and have delta derivative at the point
0
, then the composite function
F() = f((), ()) for T
has a delta derivative at that point which is expressed by the formula
F

(
0
) =
f(t
0
, s
0
)

1
t

(
0
) +
f(
1
(t
0
), s
0
)

2
s

(
0
).
Proof The proof can be seen in the reference [7].
Theorem 2.2 Let the function f be
1
-completely delta dierentiable at the point (t
0
, s
0
). If
the functions and have rst order partial delta derivative at the point (
0
,
0
), then the
composite function
F(, ) = f((, ), (, )) for (, ) T
(1)
xT
(2)
has the rst order partial delta derivatives at (
0
,
0
) which are expressed by the formulas
F(
0
,
0
)

(1)

=
f(t
0
, s
0
)

1
t
(
0
,
0
)

(1)

+
f(
1
(t
0
), s
0
)

2
s
(
0
,
0
)

(1)

and
F(
0
,
0
)

(2)

=
f(t
0
, s
0
)

1
t
(
0
,
0
)

(2)

+
f(
1
(t
0
), s
0
)

2
s
(
0
,
0
)

(2)

Proof The proof can be also seen in the reference [7].


3. The directional derivative
Let T be a time scale with the forward jump operator and the delta operator . We will
assume that 0 T. Further, let w = (w
1
, w
2
) R
2
be a unit vector and let (t
0
, s
0
) be a xed
point in R
2
. Let us set
T
1
= {t = t
0
+w
1
: T} and T
2
= {s = s
0
+w
2
: T}.
Then T
1
and T
2
are time scales and t
0
T
1
, s
0
T
2
. Denote the forward jump operators of T
1
and the delta operators by
1
.
Tangent Space and Derivative Mapping on Time Scale 5
Denition 3.1 Let a function f : T
1
xT
2
R be given. The directional delta derivative of
the function f at the point (t
0
, s
0
) in the direction of the vector w (along w) is dened as the
number
f(t
0
, s
0
)
w
= F

(0)
provided it exists, where
F() = f(t
0
+w
1
, s
0
+w
2
) for T.
Theorem 3.1 Suppose that the function f is
1
-completely delta dierentiable at the point
(t
0
, s
0
). Then the directional delta derivative of f at (t
0
, s
0
) in the direction of the w exists and
is expressed by the formula
f(t
0
, s
0
)
w
=
f(t
0
, s
0
)

1
t
w
1
+
f(
1
(t
0
), s
0
)

2
s
w
2
.
Proof The proof can be found in the reference [7].
4. The tangent vector in
n
and some properties
Let us consider the Cartesian product

n
= T
1
xT
2
x xT
n
= {P = (x
1
, x
2
, , x
n
) for x
i
T
i
}
where T
i
are dened time scale for all 1 i n, n N. We call
n
an n-dimensional Euclidean
space on time scale.
Let x
i
:
n
T
i
be Euclidean coordinate functions on time scale for all 1 i n, n N,
denoted by the set {x
1
, x
2
, , x
n
}. Let f :
n

m
be a function described by f(P) =
(f
1
(P), f
2
(P), , f
m
(P)) at a point P
n
. The function f is called
1
-completely delta
dierentiable function at the point P provided that, all f
i
, i = 1, 2, , m functions are
1
-
completely delta dierentiable at the point P. All this kind of functions set will denoted by
C

1
.
Let P
n
and {(P, v) = v
P
, P
n
} be the set of tangent vectors at the point P
denoted by V
P
(
n
). Now, we nd following properties on this set.
Theorem 4.1 Let a, b R, f, g C

1
and v
P
, w
P
, z
P
V
P
(
2
). Then, the following properties
are proven on the directional derivative.
(i)
f(t
0
, s
0
)
(av
P
+bw
P
)
= a
f(t
0
, s
0
)
v
P
+b
f(t
0
, s
0
)
w
P
,
(ii)
(af +bg)(t
0
, s
0
)
v
P
= a
f(t
0
, s
0
)
v
P
+b
g(t
0
, s
0
)
v
P
,
6 Emin

OZYILMAZ
(iii)
(fg)(t
0
, s
0
)
v
P
= g(
1
(t
0
), s
0
)
f(t
0
, s
0
)
v
P
+f(
1
(t
0
),
2
(s
0
))
g(t
0
, s
0
)
v
P

1
(t
0
)
f(t
0
, s
0
)

1
t
g(t
0
, s
0
)

1
t
v
1

2
(s
0
)
g(t
0
, s
0
)

1
t
f(
1
(t
0
), s
0
)

2
s
v
1
.
Proof Considering Denition 3.1 and Theorem 3.1, we get easily (i) and (ii). Then by
Theorem 3.1, we have
(fg)(t
0
, s
0
)
v
P
=
(fg)(t
0
, s
0
)

1
t
v
1
+
(fg)(
1
(t
0
), s
0
)

2
s
v
2
=
_
f(t
0
, s
0
)

1
t
g(t
0
, s
0
) +f(
1
(t
0
), s
0
)
g(t
0
, s
0
)

1
t
_
v
1
+
_
f(
1
(t
0
), s
0
)

2
s
g(
1
(t
0
), s
0
) +f(
1
(t
0
),
2
(s
0
))
g(
1
(t
0
), s
0
)

2
s
_
v
2
=
f(t
0
, s
0
)

1
t
g(t
0
, s
0
)v
1
+
f(t
0
, s
0
)

1
t
g(
1
(t
0
), s
0
)v
1

f(t
0
, s
0
)

1
t
g(
1
(t
0
), s
0
)v
1
+
f(
1
(t
0
), s
0
)

2
s
g(
1
(t
0
), s
0
)v
2
+f(
1
(t
0
), s
0
)
g(t
0
, s
0
)

1
t
v
1
+f(
1
(t
0
),
2
(s
0
))
g(t
0
, s
0
)

1
t
v
1
f(
1
(t
0
),
2
(s
0
))
g(t
0
, s
0
)

1
t
v
1
+f(
1
(t
0
),
2
(s
0
))
g(
1
(t
0
), s
0
)

2
s
v
2
= g(
1
(t
0
), s
0
)
_
f(t
0
, s
0
)

1
t
v
1
+
f(
1
(t
0
), s
0
)

2
s
v
2
_
+
f(t
0
, s
0
)

1
t
v
1
(g(t
0
, s
0
) g(
1
(t
0
), s
0
))
+f(
1
(t
0
),
2
(s
0
))
_
g(t
0
, s
0
)

1
t
v
1
+
g(
1
(t
0
), s
0
)

2
s
v
2
_

g(t
0
, s
0
)

1
t
v
1
(f(
1
(t
0
),
2
(s
0
)) f(
1
(t
0
), s
0
))
= g(
1
(t
0
), s
0
)
_
f(t
0
, s
0
)
v
P
_
+f(
1
(t
0
),
2
(s
0
))
_
g(t
0
, s
0
)
v
P
_

1
(t
0
)
f(t
0
, s
0
)

1
t
g(t
0
, s
0
)

1
t
v
1

2
(s
0
)
g(t
0
, s
0
)

1
t
f(
1
(t
0
), s
0
)

2
s
v
1

5. The parameter mapping and delta derivative of vector eld
along a regular curve
Denition 5.1 A -regular curve (or an arc of a -regular curve) f is dened as a mapping
x
1
= f
1
(t), x
2
= f
2
(t), , x
n
= f
n
(t), t [a, b]
Tangent Space and Derivative Mapping on Time Scale 7
of the segment [a, b] T, a < b, to the space R
n
, where f
1
, f
2
, , f
n
are real-valued functions
dened on [a, b] that are -dierentiable on [a, b]
k
with rd-continuous -derivatives and
|f

1
(t)|
2
+ |f

2
(t)|
2
+ + |f

n
(t)|
2
= 0, t [a, b]
k
.
Denition 5.2 Let f : T
n
be a -dierentiable regular curve on [a, b]
k
. Let T and T be
a time scales. A parameter mapping h : T T of the curve f is dened as t = h(s) when h
and h
1
are -dierentiable functions.
Thus, according to this new parameter, f can be written as follows:
g : T
n
, g(s) = f(h(s)).
Theorem 5.1 Let f : T
n
be a -dierentiable regular curve. Let the function h : T T
be parameter map of f and be g = f h. The -derivative of g is expressed by the formula
g

(s) =
df(h(s))

t
h

(s).
Proof Let f(t) = (f
1
(t), f
2
(t), , f
n
(t)) be a regular curve given by the vectorial form in
Euclidean space
n
. Let h be a parameter mapping of f. Considering chain rule for all f
i
, we
get
g

(s) = (
d(f
1
h)
(s)
,
d(f
2
h)
(s)
, ,
d(f
n
h)
(s)
)
= (
df
1
(h(s))

t
h

(s),
df
2
(h(s))

t
h

(s), ,
df
n
(h(s))

t
h

(s))
= (
df
1
(h(s))

t
,
df
2
(h(s))

t
, ,
df
n
(h(s))

t
)h

(s)
=
df(h(s))

t
h

(s).
Denition 5.3 A vector eld Z is a function which is associated a tangent vector to each point
of
n
and so Z(P) belongs to the set of tangent vector space V
P
(
n
) at the point P. Generally,
a vector eld is denoted by
Z(P) =
n

i=1
g
i
(t)

x
i
|
P
where g
i
(t) are real valued functions dened on T = [a, b], that are -dierentiable on [a, b]
k
with rd-continuous -derivative.
Let a function f : T
n
, f(t) = (f
1
(t), f
2
(t), , f
n
(t)) be a -dierentiable curve and
Z(P) be a vector eld along it. Thus, we dene -derivative of vector elds as follows:
8 Emin

OZYILMAZ
Denition 5.4 Let Z(P) =
n

i=1
g
i
(t)

xi
|
P
be a vector eld given along the curve f. The
-derivative of the function with respect to the parameter is dened as
Z

(P) =
dZ
t
|
P
=
n

i=1
lim
st
g
i
((t)) g
i
(s)
(t) s

x
i
|
P
=
n

i=1
g

x
i
|
P
.
Theorem 5.2 Let f be a curve given by f : T
n
and h : T T be a parameter mapping
of f. Let Z(P) =
n

i=1
g
i
(t)

xi
|
Pf(t)
be a vector eld given along the curve f. The -derivative
of the function Z(P) according to the new parameter s can be written as:
_
dZ
s
_
=
_
dZ

s
__
dh
s
_
.
Proof When we consider the chain rule on the real valued function for all the -dierentiable
functions g
i
(t), we prove Theorem 5.2.
Denition 5.5 Let a vector eld Z(P) =
n

i=1
g
i
(t)

xi
|
Pf(t)
be given along a curve f. If
dZ
t
= 0, Z(P) is called constant vector eld along the curve f.
Theorem 5.3 Let Y and Z be two vector elds along the curve f(t) and h : T R be a
-dierentiable function. Then,
(i) (Y +Z)

(t) = (Y )

(t) + (Z)

(t);
(ii) (hZ)

(t) = (h((t)))(Z)

(t) +h

(t)Z(t);
(iii) Y, Z

(t) =

Y

(t), Z((t))
_
+

Y (t), Z

(t)
_
.
where , is the inner product between the vector elds Y and Z.
Proof The (i) is obvious. Let Y (t) =
n

i=1
k
i
(t)

xi
|
f(t)
and Z(t) =
n

i=1
g
i
(t)

xi
|
f(t)
be two
vector elds. Then
(hZ)

(t) =
n

i=1
lim
st
(hg
i
)((t)) (hg
i
)(s)
(t) s

x
i
|
f(t)
=
n

i=1
lim
st
h((t))g
i
((t)) h(s)g
i
(s)
(t) s

x
i
|
f(t)
=
n

i=1
lim
st
h((t))g
i
((t)) h((t))g
i
(s) +h((t))g
i
(s) h(s)g
i
(s)
(t) s

x
i
|
f(t)
=
n

i=1
lim
st
h((t))[g
i
((t)) g
i
(s)] + [h((t)) h(s)]g
i
(s)
(t) s

x
i
|
f(t)
= h((t))
n

i=1
lim
st
g
i
((t)) g
i
(s)
(t) s

x
i
|
f(t)
+ lim
st
h((t)) h(s)
(t) s
n

i=1
g
i
(s)

x
i
|
f(t)
= (h((t)))(Z)

(t) +h

(t)Z(t).
Tangent Space and Derivative Mapping on Time Scale 9
That is the formula (ii). For (iii), we have
Y, Z

(t) =
n

i=1
lim
st
(k
i
g
i
)((t)) (k
i
g
i
)(s)
(t) s

x
i
|
f(t)
=
n

i=1
lim
st
k
i
((t))g
i
((t)) k
i
(s)g
i
(s)
(t) s

x
i
|
f(t)
=
n

i=1
lim
st
k
i
((t))g
i
((t)) k
i
(s)g
i
((t)) +k
i
(s)g
i
((t)) k
i
(s)g
i
(s)
(t) s

x
i
|
f(t)
=
n

i=1
lim
st
k
i
((t))g
i
((t)) k
i
(s)g
i
((t))
(t) s

x
i
|
f(t)
+
n

i=1
lim
st
k
i
(s)g
i
((t)) k
i
(s)g
i
(s)
(t) s

x
i
|
f(t)
=
n

i=1
lim
st
k
i
((t)) k
i
(s)
(t) s
g
i
((t)) +
n

i=1
lim
st
k
i
(s) lim
st
g
i
((t)) g
i
(s)
(t) s
=

Y

(t), Z((t))
_
+

Y (t), Z

(t)
_
.
This completes the proof.
Denition 5.6 Let f :
2

m
, f(x
1
, x
2
) = (f
1
(x
1
, x
2
), f
2
(x
1
, x
2
), , f
n
(x
1
, x
2
)) be a

1
-completely delta dierentiable function at the point P(t
0
, s
0
)
2
. For any tangent vector
v
p
V
P
(
2
), the -derivative mapping at the point P(t
0
, s
0
) of f is dened by
f

P
(v
P
) = (
f
1
(x
1
, x
2
)
v
p
,
f
2
(x
1
, x
2
)
v
p
, ,
f
m
(x
1
, x
2
)
v
p
)
f(x1,x2)
.
f

P
(v
P
) is a function from the tangent vector space V
P
(
2
) to tangent vector space V
P
(
m
).
Theorem 5.4 The function f
P
is linear mapping.
Proof Let us prove the linearity for any a R and for any two tangent vectors v
P
, w
P

V
P
(
2
). In fact,
f

P
(av
P
+bw
P
) =
_
f
1
(x
1
, x
2
)
(av
p
+bw
P
)
,
f
2
(x
1
, x
2
)
(av
p
+bw
P
)
, ,
f
m
(x
1
, x
2
)
(av
p
+bw
P
)
_
f(x1,x2)
=
_
a
f
1
(x
1
, x
2
)
v
p
+b
f
1
(x
1
, x
2
)
w
P
, , a
f
m
(x
1
, x
2
)
v
p
+b
f
m
(x
1
, x
2
)
w
P
_
f(x1,x2)
= a
_
f
1
(x
1
, x
2
)
v
p
, ,
f
m
(x
1
, x
2
)
v
p
_
f(x1,x2)
+b
_
f
1
(x
1
, x
2
)
w
P
, ,
f
m
(x
1
, x
2
)
w
P
_
f(x1,x2)
= af

P
(v
P
) +bf

P
(w
P
).
Thus the proof is completed.
10 Emin

OZYILMAZ
Denition 5.7 Let f(x
1
, x
2
) = (f
1
(x
1
, x
2
), f
2
(x
1
, x
2
), , f
n
(x
1
, x
2
)) be a
1
-completely delta
dierentiable function at the point P(x
1
, x
2
)
2
. The Jacobian matrix of f is dened by
J(f, P) =
_

_
f1(x1,x2)
1x1
f1(1(x1),x2)
1x1
f2(x1,x2)
1x1
f2(1(x1),x2)
1x1

fm(x1,x2)
1x1
fm(1(x1),x2)
1x1
_

_
m2
Theorem 5.5 Let f :
2

n
be a
1
-completely delta dierentiable function. The -
derivative mapping for any P
2
and w
P
=
_
_
w
1
w
2
_
_
V
P
(
2
) is expressed by the formula
f

P
(w
P
) = (J(f, P), w
P
)
T
.
Proof Theorem 5.5 is proven considering Denitions 5.6, 5.7 and Theorem 3.1.
References
[1] Aulbach B. and Hilger S., Linear dynamic processes with inhomogeneous time scale, In
Nonlinear Dynamics and Quantum Dynamical Systems, Berlin, Akademie Verlag, 1990,
page 9 -20.
[2] S.Hilger, Analysis on measure chains-a unied approach to continuous and discrete calculus,
Results Math.. 18(1990), 18-56.
[3] C.D.Ahlbrandt, M.Bohner and J.Ridenhour, Hamiltonian systems on time scales, J.Math.
Anal. Appl., 250(2000),561-578.
[4] J.Hoacker, Basic partial dynamic equations on time scales, J.Dierence Equ.Appl.,8(4)
(2002), 307-319(in honor of Professor Lynn Erbe).
[5] M.Bohner and A.Peterson, Dynamic Equations on Time Scales- An Introduction with Ap-
plications, Boston, Birkhauser, 2001.
[6] B.ONeill. Semi-Riemannian Geometry With Application to Relativity, Volume 1, Aca-
demic Press, New York, 1983.
[7] M.Bohner and G.Guseinov, Partial Dierentiation On Time scales, Dynamic Systems and
Applications, 12(2003), 351-379.
[8] E.

Ozyilmaz, Directional Derivative Of Vector Field And Regular Curves On Time Scales,
Applied Mathematics and Mechanics 27(10)(2006), 1349-1360.