Вы находитесь на странице: 1из 27

CHAPTER IV

Two New Subclasses of Meromorphically


Multivalent Functions Associated with the
Generalized Hypergeometric Function
4.1. Introduction
Let J denote the class of functions f of the form:
f(z) = z
p
+

k=1
a
kp
z
kp
, (4.1)
which are analytic in the punctured open unit disk U = {z C : 0 < |z| < 1}.
For two functions f(z) and g(z) given by
f(z) = z
p
+

k=1
a
kp
z
kp
and g(z) = z
p
+

k=1
b
kp
z
kp
,
we denote by (f g)(z) the Hadamard product (or convolution) of f(z) and g(z),
dene by
(f g)(z) := z
p
+

k=1
a
kp
b
kp
z
kp
=: (g f)(z).
Let f(z) and g(z) be analytic in U. We say that the function f(z) is subordinate
to g(z) in U, and we write f(z) g(z), if there exists an analytic function (z)
in U such that
|(z)| |z|, f(z) g((z)), z U.
If g(z) is univalent in U, then the following equivalent relationship holds:
f(z) g(z) f(0) = g(0), f(U) g(U).
For complex parameters
1
, ...,
q
and
1
, ...,
s

j
C/Z

0
; Z

0
:= {0, 1, 2, ...} ; j = 1, ..., s
1
we now dene the generalized hypergeometric function
q
F
s
(
1
, ...,
q
;
1
, ...,
s
; z)
[44, 53] as follows:
q
F
s
(
1
, ...,
q
;
1
, ...,
q
; z) :=

k=0
(
1
)
k
, ...(
q
)
k
(
1
)
k
, ..., (
s
)
k
z
k
k!
,
q s + 1; q, s N
0
:= N {0} ; N := {1, 2, ...} ; z U,
where ()
k
is the Pochhammer symbol (or the shifted factorial) dened (in terms
of the Gamma function) by
()
k
:=
( + k)
()
=
_

_
1 (k = 0; = 0)
( + 1)...( + k 1) (k N = {1, 2, ...} ; C).
Corresponding to a function F(
1
, ...,
q
;
1
, ...,
s
; z) dened by
F(
1
, ...,
q
;
1
, ...,
s
; z) = z
p
q
F
s
(
1
, ...,
q
;
1
, ...,
s
; z), (4.2)
Liu and Srivastava [33] considered a linear operator
H(
1
, ...,
q
;
1
, ...,
s
) : J J dened by the following Hadamard product (or
convolution):
H(
1
, ...,
q
;
1
, ...,
s
)f(z) := F(
1
, ...,
q
;
1
, ...,
s
; z) f(z). (4.3)
We note that the linear operator H(
1
, ...,
q
;
1
, ...,
s
) was motivated essen-
tially by Dziok and Srivastava [14]. Some interesting developments associated
with the generalized hypergeometric function were considered recently by Dziok
and Srivastava [15, 16] and Liu and Srivastava [34, 35]. Some authors studied
other problems of subclasses of analytic functions [41, 50, 58]. Corresponding
to the function F(
1
, ...,
q
;
1
, ...,
s
; z) dened by (4.2), we introduce a function
F

(
1
, ...,
q
;
1
, ...,
s
; z) as
F(
1
, ...,
q
;
1
, ...,
s
; z) F

(
1
, ...,
q
;
1
, ...,
s
; z) =
1
z
p
(1 z)

, > 0. (4.4)
2
Analogous to H(
1
, ...,
q
;
1
, ...,
s
) dened by (4.3), we now dene the linear
operator H
,p
(
1
, ...,
q
;
1
, ...,
s
), which was introduced by N.E. Cho and I.H.Kim
[8], as follows:
H
,p
(
1
, ...,
q
;
1
, ...,
s
)f(z) := F

(
1
, ...,
q
;
1
, ...,
s
; z) f(z),

i
,
j
C/Z

0
; i = 1, 2, ..., q; j = 1, 2, ..., s; > 0; z U; f J.
It is said that
H
,p
(
1
, ...,
q
;
1
, ...,
s
)f(z) := z
p
+

k=1
(
1
)
k
...(
s
)
k
(
1
)
k
...(
q
)
k
()
k
a
kp
z
kp
. (4.5)
For convenience, we write
H
,p
(
1
)f(z) := H
,p
(
1
, ...,
q
;
1
, ...,
s
)f(z).
It is easily veried from the denition (4.4) and (4.5) that
z(H
,p
(
1
)f(z))

= H
+1,p
(a
1
)f(z) ( + p)H
,p
(
1
)f(z), (4.6)
and
z(H
,p
(
1
)f(z))
(n)
= (H
+1,p
(
1
)f(z))
(n1)
( + p + n 1)(H
,p
(
1
)f(z))
(n1)
,(4.7)
where n N. We note that the operator H
,p
(
1
) is closely related to the Choi-
Saigo -Srivastava operator [10] for analytic functions, which includes the integral
operator studied by Liu [31] and Noor et al. [39, 40].
Now, by using the operator H
,p
(
1
), we introduce a new subclass M
,p
(; ; A, B)
such that the following subclass of meromorphic p-valent functions for f J, 0
< p, 0 1, 1 B < A 1 and
1
p
__
z
2
(H
,p
(
1
)f(z))

+ z(H
,p
(
1
)f(z))

z(H
,p
(
1
)f(z))

+ (1 )H
,p
(
1
)f(z)
_
+
_

1 + Az
1 + Bz
. (4.8)
3
Furthermore, we say that a function f M
,p
(; ; A, B) is in the analogous
subclass M
+
,p
(; ; A, B) whenever f(z) is of the form (cf.equation(4.1))
f(z) = z
p
=

k=p
a
k
z
k
, a
k
0. (4.9)
By specializing the parameters
j
(j = 1, 2, ..., q),
j
(j = 1, 2, ..., s), , , A, B and
p in the subclass M
,p
(; ; A, B), See([8, 9, 34, 59]).
(1) M
,1
(0; ; A, B) = MS
,1
(q, s, ; , A, B) and
M
,
1
(1; ; A, B) = MK
,1
(q, s, ; , A, B) (Cho and Kim [8]).
(2) If H
,p
(
1
)f(z) = H
,1
(a, 1; b)f(z) = I

(a, b)f(z), then


M
,1
(0; ; A, B) = MS

a,b
(; A, B) and M
,1
(1; ; A, B) = MK

a,b
(; A, B)
(Cho and Noor [9]);
(3) If H
,p
(
1
)f(z) = H
,1
(n + 1, a; a)f(z) = I
n,
f(z), then
M
,1
(0; ; 1, 1) = S
n,
(1; ; 1, 1) = C

n,
() (Yuan et al [59]);
(4) If H
,p
(
1
)f(z) = H
,p
(, c; a)f(z) = L
p
(a, c)f(z), then
M
,p
(0; 0; A, B) = H
a,c
(p; A, B) and M
+
,p
(0; 0; A, B) = H
+
a,c
(p, A, B)
(Liu and Srivastava [34]).
Furthur, we write H
,p
f(z) = H
,p
(
1
)f(z).
In this chapter, we investigate the various properties and characteristics of
the subclass M
,p
(; ; A, B) and M
+
,p
(; ; A, B). Following the recent inves-
tigate by Altintas et al. [3], we extend the concept of neighborhoods of ana-
lytic functions, which was considered earlier by (for example) Goodman [25] and
Ruscheweyh [48], to meromorphically multivalent functions belonging to the class
M
,p
(; ; A, B) and M
+
,p
(; ; A, B). We also derive many interesting results
on the Hadamard products of functions belonging to the p-valently meromorphic
function class M
+
,p
(; ; A, B).
4
We begin by recalling the following result (popularly known as Jacks Lemma),
which we shall apply in proving our rst inclusion theorem (Theorem 4.1 below).
Jacks Lemma [26]. Let the (nonconstant) function w(z) be analytic in U with
w(0) = 0. If |w(z)| attains its maximum value on the circle |z| = r < 1 at a point
z
0
U, then z
0
w

(z
0
) = w(z
0
), where is a real number and 1.
4.2. Inclusion Properties of the Class M
,p
(; ; A, B)
We rst prove the following inclusion relation.
Theorem 4.1. If

(p )(A B)
1 + B
, 1 < B < A 1; p N,
then
M
+1,p
(; ; A, B) M
,p
(; ; A, B).
Let f M
+1,p
(; ; A, B) and suppose that
1
p
__
z
2
(H
,p
f(z))

+ z(H
,p
f(z))

z(H
,p
f(z))

+ (1 )H
,p
f(z)
_
+
_

1 + Aw(z)
1 + Bw(z)
. (4.10)
It is said that
z
2
(H
,p
f(z))

+ z(H
,p
f(z))

z(H
,p
f(z))

+ (1 )H
,p
f(z)
+ + p
=
+ [B (p )(A B)]w(z)
1 + Bw(z)
, (4.11)
where the function w(z) is either analytic or meromorphic in U, with w(0) = 0.
By dierentiating both sides of (4.11) with respect to z logarithmically and using
the identity (4.6) and (4.7), we have
z
2
(H
+1,p
f(z))

+ z(H
+1,p
f(z))

z(H
+1,p
f(z))

+ (1 )H
+1,p
f(z)
5
=
z
3
(H
,p
f(z))

+ ( + p + 2 + 1)z
2
(H
,p
f(z))

+ ( + p + 1)z(H
,p
f(z))

z
2
(H
,p
f(z))

+ ( + p + 1)z(H
,p
f(z))

+ ( + p)(1 )H
,p
f(z)
=
(p )(A B)zw

(z)
[1 + Bw(z)][ + (B (p )(A B))w(z)]

p + [pA (A B)]w(z)
1 + Bw(z)
.
(4.12)
We claim that
|w(z)| < 1, z U.
Otherwise, there exists a point z
0
U such that
max
|z||z
0
|
|w(z
0
)| = 1.
Applying Jacks Lemma, we have
z
0
w

(z
0
) = w(z
0
), 1.
Now, upon setting
w(z
0
) = e
i
, 0 2.
If we put z = z
0
in (4.12), we get
_
_
_
_
_

z
2
(H
+1,p
f(z))

+z(H
+1,p
f(z))

z(H
+1,p
f(z))

+(1)H
+1,p
f(z)
+ p
B
_
z
2
(H
+1,p
f(z))

+z(H
+1,p
f(z))

z(H
+1,p
f(z))

+(1)H
+1,p
f(z)++p
_
(p )(A B)

z=z
0
_
_
_
_
_
2
1
=

(p)(AB)e
i
[1+Be
i
][+(B(p)(AB))e
i
]

p+[pA(AB)]e
i
1+Be
i
+ p
B
_
(p)(AB)e
i
[1+Be
i
][+(B(p)(AB))e
i
]

p+[pA(AB)]e
i
1+Be
i
+ p
_
(p )(A B)

2
1
=

( + ) + [B (p )(A B)]e
i
[( )B + (p )(A B)]e
i

1 (4.13)
=
[ + 2 + (2 )B
2
2B(p )(A B)] + 2[2B (p )(A B)] cos
| [( )B + (p )(A B)]e
i
|
2
.
6
Set
g(t) = [ + 2 + (2 )B
2
2B(p )(A B)][2B (p )(A B)]t.
Then, by conditions, we have
g(1) = (1 + B)[2(1 + B) 2(p )(A B) + (1 B)] > 0,
and
g(1) = (1 B)[2(1 B) + 2(p )(A B) + (1 + B)] > 0,
which, together, imply that
g(cos ) 0, 0 2. (4.14)
In view of(4.13) and (4.14) it would obviously contradict our hypothesis that
f M
+1,p
(; ; A, B).
Hence, we must have
|w(z)| < 1, z U
and we conclude from (4.10) that
f M
,p
(; ; A, B).
The proof of Theorem 4.1 is thus completed. Next, we prove an inclusion property
associated with a certain integral transform.
7
Theorem 4.2.
Let be a complex number such that
Re()
(p )(A B)
1 + B
, 1 < B < A 1; p N.
If f M
,p
(; ; A, B), then the function F(z) dened by
F(z) :=

z
+p
_
z
0
t
+p1
f(t)dt (4.15)
also belongs to the class M
,p
(; ; A, B).
Proof.
From (4.15), we readily have
z(H
,p
F(z))

= H
,p
f(z) ( + p)H
,p
F(z), (4.16)
and
z(H
,p
F(z))

= (H
,p
f(z))

( + p + 1)(H
,p
F(z))

. (4.17)
Put
1
p
__
z
2
(H
,p
F(z))

+ z(H
,p
F(z))

z(H
,p
F(z))

+ (1 )H
,p
F(z)
_
+
_
=
1 + Aw(z)
1 + Bw(z)
,
it is said that
z
2
(H
,p
F(z))

+ z(H
,p
F(z))

z(H
,p
F(z))

+ (1 )H
,p
F(z)
+ + p
=
+ [B (p )(A B)]w(z)
1 + Bw(z)
, (4.18)
where the function w(z) is either analytic or meromorphic in U, with w(0) = 0.
Using (4.16),(4.17) and the identity (4.18), we obtain
z
2
(H
,p
f(z))

+ z(H
,p
f(z))

z(H
,p
f(z))

+ (1 )H
,p
f(z)
8
=
(p )(A B)zw

(z)
[1 + Bw(z)][ + (B (p )(A B))w(z)]

p + [pA (A B)]w(z)
1 + Bw(z)
.
The remaining part of the proof Theorem 4.2 is similar to that of Theorem 4.1
and so is omitted.
Theorem 4.3.
The f M
,p
(; ; A, B) if and only if the function F(z) given by
F(z) =

z
+p
_
z
0
t
+p1
f(t)dt (4.19)
is in the class M
+1,p
(; ; A, B).
Proof.
By using of (4.19), we have
f(z) = ( + p)F(z) + zF

(z), (4.20)
which, in the right hand of (4.6), implies
H
,p
f(z) = ( + p)H
,p
F(z) + z(H
,p
F(z))

= H
+1,p
F(z).
Therefore, we have
H
,p
f(z) = H
+1,p
F(z),
and the desired result follows at once.
Remark 4.1.
Setting p = 1, = 0 or = 1 in Theorem 4.1 and Theorem 4.2, we get the part
results of Corollary 1 and Corollary 2 in Cho and Kim [8], which contain some
results in Cho and Noor [9] and Theorem 1 in Liu and Srivastava [34].
9
4.3. Properties of the class M
+
,p
(; ; A, B)
In this section we assume further that

j
> 0, j = 1, ..., q;
j
> 0, j = 1, ..., s and 0 B < A 1.
Theorem 4.4.
Let
1
p+1
< 1, f J be given by (4.9). Then f M
+
,p
(; ; A, B) if and
only if

k=p
[(1 + B)(k + p) + (p )(A B)]
k+p
()a
k
(p )(A B)(p + 1),
(4.21)
where

m
() = [m + 1 (p + 1)]
(
1
)
m
...(
s
)
m
(
1
)
m
...(
q
)
m
()
m
, m = 1, 2, 3, ... (4.22)
Each of these inequalities is sharp, with the extremal function f
(0)
k
(z) given by
f
(0)
k
(z) = z
p
+
(p )(A B)(p + 1)
[(1 + B)(k + p) + (p )(A B)]
k+p
()
z
k
,
k = p, p + 1, p + 2, ...; p N. (4.23)
Proof.
Suppose that the function f(z), given by (4.9), is in the class M
+
,p
(; ; A, B).
Then, form (4.5) and (4.8), we nd that

z
2
(H
,p
f(z))

+ z(H
,p
f(z))

+ p[z(H
,p
f(z))

+ (1 )H
,p
f(z)]
B[z
2
(H
,p
f(z))

+ z(H
,p
f(z))

] + [Bp + (p )(A B)][z(H


,p
f(z))

+ (1 )H
,p
f(z)]

k=p
(k + p)
k+p
()a
k
z
k+p
(p )(A B)(p + 1)

k=p
[B(k + p) + (p )(A B)]
k+p
()a
k
z
k+p

< 1, z U.
10
Now, since |Re(z)| |z|, we have
Re
_

k=p
(k + p)
k+p
()a
k
z
k+p
(p )(A B)(p + 1)

k=p
[B(k + p) + (p )(A B)]
k+p
()a
k
z
k+p
_
< 1.
(4.24)
We choose the values of z on the real axis with z = r, 0 r < 1. Then, for r = 0,
the denominator of (4.24) is positive and so is positive for all r(0 < r < 1) by
k +p > B(k +p) +(p )(AB) (k = p, p +1, ...; p N). Letting z = r 1

,
(4.24) yields

k=p
(k + p)
k+p
()a
k
(p )(A B)(p + 1)

k=p
[B(k + p) + (p )(A B)]
k+p
()a
k
,
which leads us to the desired inequality (4.21).
Conversely, by applying the hypothesis (4.21), we get

z
2
(H
,p
f(z))

+ z(H
,p
f(z))

+ p[z(H
,p
f(z))

+ (1 )H
,p
f(z)]
B[z
2
(H
,p
f(z))

+ z(H
,p
f(z))

] + [Bp + (p )(A B)][z(H


,p
f(z))

+ (1 )H
,p
f(z)]

k=p
(k + p)
k+p
()a
k
(p )(A B)(p + 1)

k=p
[B(k + p) + (p )(A B)]
k+p
()a
k

1, 0 B < A 1; z C; |z| = 1.
Hence, by the maximum modulus theorem, we have f(z) M
+

(; ; A, B).
By observing that the functions f
(0)
k
(z) given by (4.23) are indeed the extremal
functions for the assertion (4.21), we complete the proof of Theorem 4.4.
By applying Theorem 4.4, we obtain the following sharp coecient estimates.
11
Corollary 4.1.
Let
1
p+1
< 1, f J be given by (4.9). If f M
+
,p
(; ; A, B), then
a
k

(p )(A B)(p + 1)
[(1 + B)(k + p) + (p )(A B)]
k+p
()
,
k = p, p + 1, p + 2, ...; p N. (4.25)
Each of these ineqalities is sharp, with the extremal function f
(0)
k
(z) given by
(4.23). Next we prove the following growth and distortion properties for the class
M
+
,p
(; ; A; B).
Theorem 4.5.
Let
1
p+1
< 1 and f M
+
,p
(; ; A, B). If the sequence {C
k
} is nonde-
creasing, then
r
p

(p )(A B)(p + 1)
C
p
r
p
|f(z)|
r
p
+
(p )(A B)(p + 1)
C
p
r
p
,
0 < |z| = r < 1, (4.26)
where
C
k
= [(1 + B)(k + p) + (p )(A B)]
k+p
(),
k = p, p + 1, p + 2, ...; p N, (4.27)
and
m
() is given by Theorem 4.4.
If the sequence {C
k
/k} is nondecreasing, then
pr
p1

(p )(A B)(p + 1)
C
p
r
p1
|f

(z)|
pr
p1
+
(p )(A B)(p + 1)
C
p
r
p1
,
0 < |z| = r < 1. (4.28)
12
Each of these results is sharp, with the extremal functions f
(0)
k
(z) given by (4.23).
Proof.
Let the function f(z), given by (4.9), be in the class M
+
,p
(; ; A, B). If the
{C
k
} is nondecreasing, then (by Theorem 4.4) we have

k=p
a
k

(p )(A B)(p + 1)
C
p
. (4.29)
On the other hand, if the sequence {C
k
/k} is nondecreasing, Theorem 4.4 also
implies

k=p
ka
k
p
(p )(A B)(p + 1)
C
p
. (4.30)
Thus, assertions (4.26) and (4.28) follow immediately.
Finally, it is easy to see that the bounds in (4.26) and (4.28) are attained for the
function f(z) given by (4.23) with k = p.
Next we determine the radii of meromorphically p-valent starlikeness of order
(0 < p) and meromorphically p-valent convexity of order (0 < p) for
functions in the class M
+
,p
(; ; A, B).
Theorem 4.6.
Let the function f(z) dened by (4.9) be in the class M
+
,p
(; ; A, B). Then
we have:
(i) f(z) is meromorphically p-valent starlike of order (0 < p) in the disc
|z| < 1, that is,
Re
_

zf

(z)
f(z)
_
> , |z| < r
1
; 0 < p; p N, (4.31)
where
r
1
= inf
kp
_
(p )[(1 + B)(k + p) + (p )(A B)]
k+p
()
(k + )(p )(A B)(p + 1)
_ 1
k+p
. (4.32)
13
(ii) f(z) is meromorphically p-valent convex of order (0 < p) in the disc
|z| < 1, that is,
Re
_

_
1 +
zf

(z)
f

(z)
__
> , |z| < r
2
; 0 < p; p N, (4.33)
where
r
2
= inf
kp
_
p(p )[(1 + B)(k + p) + (p )(A B)]
k+p
()
k(k + )(p )(A B)(p + 1)
_ 1
k+p
. (4.34)
Each of these results is sharp, with the extremal functions f
(0)
k
(z) given by (4.23).
Proof.
(i). From the denition (4.9), we easliy get

zf

(z)
f(z)
+ p
zf

(z)
f(z)
p + 2

k=p
(k + p)a
k
|z|
k+p
2(p )

k=p
(k p + 2)a
k
|z|
k+p
. (4.35)
Thus, we have the desired inequality

zf

(z)
f(z)
+ p
zf

(z)
f(z)
p + 2

1, 0 < p; p N, (4.36)
if

k=p
_
k +
p
_
a
k
|z|
k+p
1. (4.37)
Hence, by Theorem 4.4, (4.37) will be true if
_
k +
p
_
|z|
k+p

_
[(1 + B)(k + p) + (p )(A B)]
k+p
()
(p )(A B)(p + 1)
_
,
k p; p N. (4.38)
The last inequality (4.38) leads us immediately to the disc |z| < r
1
, where r
1
is
given by (4.32).
14
(ii). In order to prove the second assertion of Theorem 4.6, we nd from the
denition (4.9) that

1 +
zf

(z)
f

(z)
+ p
1 +
zf

(z)
f

(z)
p + 2

k=p
k(k + p)a
k
|z|
k+p
2p(p )

k=p
k(k p + 2)a
k
|z|
k+p
. (4.39)
Thus, we have the desired inequality

1 +
zf

(z)
f

(z)
+ p
1 +
zf

(z)
f

(z)
p + 2

1, 0 < p; p N, (4.40)
if

k=p
k(k + )
p(p )
a
k
|z|
k+p
1. (4.41)
Hence, by Theorem 4.4, (4.41) will be true if
k(k + )
p(p )
|z|
k+p

_
[(1 + B)(k + p) + (p )(A B)]
k+p
()
(p )(A B)(p + 1)
_
,
k p; p N. (4.42)
The last inequality (4.42) readily yields the disc |z| < r
2
, where r
2
is dened by
(4.34), and the proof of Theorem 4.6 is completed by merely verifying that each
assertion is sharp for the functions f
(0)
k
(z) given by (4.23).
4.4 Neighborhoods and Partial Sums
In this section, we also assume that

j
> 0, j = 1, ..., q;
j
> 0, j = 1, ..., s; 1 B < A 1.
Following the earlier works (based upon the familiar concept of neighborhoods of
analytic functions) by Goodman [25] and Ruscheweyh [48], and (more recently)
15
by Altintas et al. [3,4], Altintas and Owe [2], Liu [30, 32] and Liu and Srivastava
[33, 34], we begin by introducing here the -neighborhood of a function f(z) J
of the form (4.1) by means of the denition given below:
N

(f) =
_

_
g J :

k=1
[(1+|B|)k+(p)(AB)]|
k
()|
(p)(AB)|p+1|
|b
kp
a
kp
| ,
and g(z) = z
p
+

k=1
b
kp
z
kp
, > 0.
(4.43)
Where
k
() is dened by (4.22).
Theorem 4.7.
Let 0 1, =
1
p+1
and f M
,p
(; ; A, B) be given by (4.1). If f(z)
satises the following condition:
f(z) + z
p
1 +
M
,p
(; ; A, B), C, || < , || < 1, (4.44)
then
N

(f) M
,p
(; ; A, B). (4.45)
Proof.
It is easily seen from (4.8) that (z) = z
p
+

k=1
d
kp
z
kp
M
,p
(; ; A, B)
if and only if for any complex number with || = 1,

z
2
(H
,p
(z))

+ z(H
,p
(z))

+ p[z(H
,p
(z))

+ (1 )H
,p
(z)]
B[z
2
(H
,p
(z))

+ z(H
,p
(z))

] + [(p )(A B) + Bp][z(H


,p
(z))

+ (1 )H
,p
(z)]

k=1
k
k
()d
kp
z
k
(p )(A B)(p + 1 ) +

k=1
[Bk + (p )(A B)]
k
()d
kp
z
k

= , z U, (4.46)
which is equivalent to
( h)(z)
z
p
= 0, z U, (4.47)
16
where, for convenience,
h(z) = z
p
+

k=1
c
k
pz
kp
(4.48)
= z
p
+

k=1
[(1 B)k (p )(A B)]
k
()
(p )(A B)(p + 1)
z
kp
.
Form (4.48), we have
|c
kp
| =

[(1 B)k (p )(A B)]


k
()
(p )(A B)(p + 1)

[(1 +|B|)k + (p )(A B)]|


k
()|
(p )(A B)|p + 1|
, k, p N.
Now, if f(z) = z
p
+

k=1
a
kp
z
kp
J satises the condition (4.44), then (4.47)
yields

(f h)(z)
z
p

, z U; > 0. (4.49)
By letting
g(z) = z
p
+

k=1
b
kp
z
kp
N

(f), (4.50)
we have

[g(z) f(z)] h(z)


z
p

k=1
(b
kp
a
kp
)c
kp
z
k

|z|

k=1
[(1 +|B|)k + (p )(A B)]|
k
()|
(p )(A B)|p + 1|
|b
kp
a
kp
|
< , z U; > 0.
Thus we have (4.47), and we also have (4.46) for any C such that || = 1,
which implies that g(z) M
,p
(; ; A, B). This evidently proves the assertion
17
(4.46) of Theorem 4.7.
Remark 4.2.
For = 0, = 0 and H
,p
(
1
)f(z) = H
,p
(, c, ; a)f(z) = L
p
(a, c) f(z) in
Theorem 4.7, we get Theorem 6 in Liu and Srivastava [34].
We now dene the neighborhood of a function f(z) J of the form (4.9)
as follows:
N
+

(f) =
_

_
g J :

k=p
[(1+|B|)(k+p)+(p)(AB)]
k+p
()
(p)(AB)(p+1)
|b
k
a
k
|
and g(z) = z
p
+

k=p
b
k
z
k
, b
k
0; 0;
1
p+1
< 1,
(4.51)
where
k
() is dened by (4.22).
Theorem 4.8.
Let 0 B < 1, f M
+
+1,p
(; ; A; B) be given by (4.9). Then
N
+

(f) M
+
,p
(; ; A, B), :=
2p
+ 2p
. (4.52)
The result is sharp in the sense that cannot be increased.
Proof.
Making use of the same method as in the proof Theorem 4.7, we can show
that (cf. equation (4.49))
h(z) = z
p
+

k=p
c
k
z
k
= z
p
+

k=p
[(1 B)(k + p) (p )(A B)]
k+p
()
(p )(A B)(p + 1)
z
k
. (4.53)
Thus, under the hypothesis 0 B < A 1, 0 < p and p N, if f(z)
M
+
+1,p
(; ; A, B) is given by (4.9), we obtain

(f h)(z)
z
p

1 +

k=p
a
k
c
k
z
k+p

18
1

+ 2p

k=p
[(1 + B)(k + p) + (p )(A B)]
k+p
( + 1)
(p )(A B)(p + 1)
a
k
.
Also, from Theorem 4.4, we obtain

(f h)(z)
z
p

1

+ 2p
=
2p
+ 2p
=
The remaining part of the proof of Theorem 4.8 is similar to that of Theorem 4.7,
and we skip the details involved.
To show the sharpness, we consider the function f(z) and g(z) given by
f(z) = z
p
+
(p )(A B)(p + 1)
[2p(1 + B) + (p )(A B)]
2p
( + 1)
z
p
M
+
+1,p
(; ; A, B) (4.54)
and
g(z) = z
p
+
(p )(A B)(p + 1)
[2p(1 + B) + (p )(A B)]
_
1

2p
( + 1)
+

2p
()
_
z
p
, (4.55)
where

> =
2p
+2p
. Clearly, the function g(z) belongs to N
+

(f). On the other


hand, we nd from Theorem 4.4 that g(z) is not in the class M
+
,p
(; ; A, B).
Thus the proof Theorem 4.8 is completed.
Theorem 4.9.
Let 0 B < 1 and be a real number with
>
(p )(A B)
1 + B
.
If the function f(z), given by (4.9), is in the class M
+
,p
(; ; A, B), then the func-
tion F(z) dened by (4.15) belongs to N
+
1
(f). The result is sharp in the sense
that the constant 1 cannot be decreased.
19
Proof.
Suppose that the function f(z), given by (4.9), is the class M
+
,p
(; ; A; B).
Then it follows from (4.15) and Theorem 4.2 that
F(z) = z
p
+

k=p
b
k
z
k
= z
p
+

k=p

+ p + k
a
k
z
k
M
+
,p
(; ; A, B). (4.56)
From the hypothesis of Theorem 4.9, we have

k=p
[(1 + B)(k + p) + (p )(A B)]
k+p
()
(p )(A B)(p + 1)
|b
k
a
k
|
=

k=p
[(1 + B)(k + p) + (p )(A B)]
k+p
()
(p )(A B)(p + 1)
.
p + k
+ p + k
a
k

k=p
[(1 + B)(k + p) + (p )(A B)]
k+p
()
(p )(A B)(p + 1)
a
k
1, f M
+
,p
(; ; A, B),
which shows that F(z) N
+
1
(f). In order to verify the sharpness of the assertion
Theorem 4.9, we consider the function f
(0)
k
(z) given by (4.23). From (4.23) and
(4.55), we have
F(z) =

z
+p
_
z
0
t
+p1
f
(0)
k
(t)dt
= z
p
+
(p )(A B)(p + 1)
[(1 + B)(k + p) + (p )(A B)]
k+p
()
.

+ p + k
z
k
,
k = p, p + 1, ...; p N.
Thus, by making use of (4.51), we get

k=p
[(1 + B)(k + p) + (p )(A B)]
k+p
()
(p )(A B)(p + 1)
|b
k
a
k
|
20
=
p + k
+ p + k
1, k (4.57)
which clearly shows that the constant 1 is the best possible. This evidently com-
pletes the proof of Theorem 4.9.
Our proof of Theorem 4.10 below is similar to that of Theorem 4.8 and so is
omitted.
Theorem 4.10.
Let 0 B < 1 and be a real number such that
>
(p )(A B)
1 + B
.
If f M
+
+1,p
(; ; A, B), then
N
+

(F) M
+
,p
(; ; A, B),

:=
2p(2p + + )
( + 2p)( + 2p)
, (4.58)
and F(z) dened by (4.15). The result is sharp in the sense that

cannot be
increased.
Theorem 4.11.
Let 0 1, =
1
p+1
and f J be given by (4.1), dene the partial sums
s
n
(z) by
s
1
(z) = z
p
, s
n
(z) = z
p
+
n1

k=1
a
kp
z
kp
, n N/ {1} . (4.59)
Suppose also that

k=1
d
k
|a
kp
| 1, d
k
=
[(1 +|B|)k + (p )(A B)]
k
()
|(p )(A B)|p + 1|
. (4.60)
Then
21
(i) f(z) M
,p
(; ; A, B).
(ii) If {
k
()} (k N) is nondecreasing and

1
() >
(p )(A B)|p + 1|
[(1 + B)(1 + p) + (p )(A B)]
, (4.61)
then
Re
_
f(z)
s
n
(z)
_
> 1
1
d
n
, z U; n N (4.62)
and
Re
_
s
n
(z)
f(z)
_
>
d
n
1 + d
n
, z U; n N. (4.63)
Each of the bounds in (4.62) and (4.63) is the best possible for each n N.
Proof.
(i). It is not dicult to see that
z
p
M
,p
(; ; A, B), p N. (4.64)
Thus, from Theorem 4.7 and the hypothesis (4.60) of Theorem 4.11 we have
f(z) N
1
(z
p
) M
,p
(; ; A, B)
as asserted by Theorem 4.11 (ii) Under the hypothesis in part (ii) Theorem 4.11,
we can see from (4.60) that
d
k+1
> d
k
> 1, k N. (4.65)
Therefore, we have
n1

k=1
|a
kp
| + d
n

k=n
|a
kp
|

k=1
d
k
|a
kp
| 1, (4.66)
by using hypothesis (4.60) of Theorem 4.11 again.
By setting
g
1
(z) = d
n
_
f(z)
s
n
(z)

_
1
1
d
n
_
_
= 1 +
d
n

k=n
a
kp
z
k
1 +

n1
k=1
a
kp
z
k
, (4.67)
22
and applying (4.66), we nd that

g
1
(z) 1
g
1
(z) + 1

d
n

k=n
a
kp
2 2

n1
k=1
|a
kp
| d
n

k=n
|a
kp
|
1, z U, (4.68)
which readily yields the assertion (4.62) of Theorem 4.11. If we take
f(z) = z
p

z
np
d
n
, (4.69)
then
f(z)
s
n
(z)
= 1
z
n
d
n
1
1
d
n
, z 1

,
which shows that the bound in (4.62) is the best possible for each n N.
Similarly, if we put
g
2
(z) = (1 + d
n
)
_
s
n
f(z)
f(z)

d
n
1 + d
n
_
= 1
(1 + d
n
)

k=n
a
kp
z
k
1 +

k=1
a
kp
z
k
and make use of (4.66), we can deduce that

g
2
(z) 1
g
2
(z) + 1

(1 + d
n
)

k=n
|a
kp
|
2 2

n1
k=1
|a
kp
| + (1 d
n
)

k=n
|a
kp
|
1, z U, (4.70)
which leads us immediately to the assertion (4.63) of Theorem 4.11.
The bound in (4.63) is sharp for each n N, with the extremal function f(z)
given by (4.69). The proof of Theorem 4.11 is thus completed.
Remark 4.3.
For = 0, = 0 and H
,p
(
1
)f(z) = H
,p
(, c; a)f(z) = L(a, c) f(z) in
Theorem 4.6, we get Theorem 7 in Liu and Srivastava [34].
4.5. Convolution Properties
In this section, we rst set
j
> 0, j = 1, ..., q;
j
> 0, j = 1, ..., s;
1
p+1
<
1; 0 B < A 1 and dene functions f
j
(z)(j = 1, 2) by
f
j
(z) = z
p
+

k=p
a
k,j
z
k
, a
k,j
0, j = 1, 2 (4.71)
23
and dene the Hadamard product (or convolution) of the functions f
1
(z) and f
2
(z)
as
(f
1
f
2
)(z) = z
p
+

k=p
a
k,1
a
k,2
z
k
. (4.72)
Throughout this section, we assume further that the sequence
_

k+p
()
k+p
_
(k
p; p N) is nondecreasing and
2p
() > 0.
Theorem 4.12.
Let the functions f
j
(z)(j = 1, 2) dened by (4.71) be in the class M
+
,p
(; ; A, B).
Then (f
1
f
2
)(z) M
+
,p
(; ; A, B), where
= p
2p(p )
2
(1 + B)(A B)(p + 1)
[2p(1 + B) + (p )(A B)]
2

2p
() [(p )(A B)]
2
(p + 1)
,
k p.(4.73)
The results is sharp for the functions g
j
(z)(j = 1, 2) given by
g
j
(z) = z
p
+
(p )(A B)(p + 1)
[2p(1 + B) + (p )(A B)]
2p
()
z
p
, j = 1, 2; p N. (4.74)
Proof.
Employing the techniques used earlier by Schild and Silverman [49], we need
to nd the largest such that

k=p
[(1 + B)(k + p) + (p )(A B)]
k+p
()
(p )(A B)(p + 1)
a
k,1
a
k,2
1, (4.75)
for f
j
(z) M
+
,p
(; ; A, B), j = 1, 2. Since f
j
(z) M
+
,p
(; ; A, B), j = 1, 2, we
readily see that

k=p
[(1 + B)(k + p) + (p )(A B)]
k+p
()
(p )(A B)(p + 1)
a
k,j
1, j = 1, 2. (4.76)
24
Therefore, by the Cauchy-Schwarz inequality, we obtain

k=p
[(1 + B)(k + p) + (p )(A B)]
k+p
()
(p )(A B)(p + 1)

a
k,1
a
k,2
1. (4.77)
This implies that we only need to show that
[(1 + B)(k + p) + (p )(A B)]
(p )
a
k,1
a
k,2

[(1 + B)(k + p) + (p )(A B)]


(p )

a
k,1
a
k,2
, k p, (4.78)
or, equivalently that

a
k,1
a
k,2

(p )[(1 + B)(k + p) + (p )(A B)]
(p )[(1 + B)(k + p) + (p )(A B)]
, k p. (4.79)
Hence, by inequality (4.77), it is sucient to prove that
(p )(A B)(p + 1)
[(1 + B)(k + p) + (p )(A B)]
k+p
()

(p )[(1 + B)(k + p) + (p )(A B)]


(p )[(1 + B)(k + p) + (p )(A B)]
, k p. (4.80)
It follows from (4.80) that
p
(1 + B)(p )
2
(A B)(p + 1)
[(1+B)(k+p)+(p)(AB)]
2

k+p
()
k+p

[(p)(AB)]
2
(p+1)
k+p
,
k p. (4.81)
Now, dening the function (k) by
(k) = p
(1 + B)(p )
2
(A B)(p + 1)
[(1+B)(k+p)+(p)(AB)]
2

k+p
()
k+p

[(p)(AB)]
2
(p+1)
k+p
, k p.
We see that (k) is an increasing function of k. Therefore, we conclude that
(p)
= p
2p(1 + B)(p )
2
(A B)(p + 1)
[2p(1 + B) + (p )(A B)]
2

2p
() [(p )(A B)]
2
(p + 1)
,
k p.
25
which evidently completes the proof of Theorem 4.12.
Using the same method as in our proof of Theorem 4.12, we obtain the following
result:
Theorem 4.13.
Let the function f
1
(z) dened by (4.71) be in the class M
+
,p
(; ; A, B), and
the function f
2
(z) dened by (4.71) be in the class M
+
,p
(; ; A, B). Then (f
1

f
2
)(z) M
+
,p
(; ; A, B), where
= p
2p(p )(p )(1 + B)(A B)(p + 1)
M N
2p
() (p )(p )(A B)
2
(p + 1)
, k p, (4.82)
M = [2p(1 + B) + (p )(A B)] and N = [2p(1 + B) + (p )(A B)].
The result is sharp for the functions f
j
(z)(j = 1, 2) given by
f
1
(z) = z
p
+
(p )(A B)(p + 1)
[2p(1 + B) + (p )(A B)]
2p
()
z
p
, p N, (4.83)
and
f
2
(z) = z
p
+
(p )(A B)(p + 1)
[2p(1 + B) + (p )(A B)]
2p
()
z
p
, p N. (4.84)
Theorem 4.14.
Let the functions f
j
(z) (j = 1, 2) dened by (4.71) be in the class M
+
,p
(; ; A, B).
Then the function h(z) dened by
h(z) = z
p
+

k=p
(a
2
k,1
+ a
2
k,2
)z
k
(4.85)
belongs to the class M
+
,p
(; ; A, B), where
= p
4p(p )
2
(1 + B)(A B)(p + 1)
[2p(1 + B) + (p )(A B)]
2

2p
() 2[(p )(A B)]
2
(p + 1)
.
26
The result is sharp for the functions g
j
(z)(j = 1, 2) given already by (4.74).
Proof.

k=p
_
[(1 + B)(k + p) + (p )(A B)]
k+p
()
(p )(A B)(p + 1)
_
2
a
2
k,j

_
_

k=p
[(1 + B)(k + p) + (p )(A B)]
k+p
()
(p )(A B)(p + 1)
a
k,j
_
_
2
1, j = 1, 2. (4.86)
For f
j
(z) M
+
,p
(; ; A, B)(j = 1, 2), we have

k=p
1
2
_
[(1 + B)(k + p) + (p )(A B)]
k+p
()
(p )(A B)(p + 1)
_
2
[a
2
k,1
+ a
2
k,2
] 1.
(4.87)
In order to obtain our result, we have to nd the largest such that
2[(1 + B)(k + p) + (p )(A B)]
p

[(1 + B)(k + p) + (p )(A B)]


2

k+p
()
(p )
2
(A B)(p + 1)
, k p.
So that
p
2(k + p)(p )
2
(1 + B)(A B)(p + 1)
[(1 + B)(k + p) + (p )(A B)]
2

k+p
() 2[(p )(A B)]
2
(p + 1)
,
k p.
Now, we dene the function (k) by
(k) =
[(1 + B)(k + p) + (p )(A B)]
2

k+p
()
k + p

2[(p )(A B)]
2
(p + 1)
k + p
,
k p.
We oberve that (k) is an increasing function of k(k p). Thus, we conclude
that
(p) = p
4p(p )
2
(1 + B)(A B)(p + 1)
[2p(1 + B) + (p )(A B)]
2

2p
() 2[(p )(A B)]
2
(p + 1)
,
which completes the proof of Theorem 4.14.
27

Вам также может понравиться