Bulletin of the Iranian Mathematical Society

Starlike Functions Associated with Cosine Functions

--Manuscript Draft--

Manuscript Number: BIMS-D-20-00022

Full Title: Starlike Functions Associated with Cosine Functions

Article Type: Original Article

Abstract: Let S_{cos}^{∗} denote the class of normalized analytic functions f such that
((zf′(z))/(f(z)))≺cos(z). For this class, we obtain structural formula, inclusion results,
differential subordinations and some sharp radii problems such as radius of convexity,
radius for the class of Janowski starlike functions and radius for some other subclasses
of starlike functions.

Powered by Editorial Manager® and ProduXion Manager® from Aries Systems Corporation
 Click here to download Manuscript 514.tex

1
2
3
4
5
6
7
8 STARLIKE FUNCTIONS ASSOCIATED WITH COSINE
9 FUNCTIONS
10
11
12 KHADIJA BANO1 , MOHSAN RAZA2,∗
13
14 Abstract. Let Scos ∗
denote the class of normalized analytic functions f such
15 zf 0 (z)
that f (z) ≺ cos(z). For this class, we obtain structural formula, inclusion re-
16
17 sults, differential subordinations and some sharp radii problems such as radius
18 of convexity, radius for the class of Janowski starlike functions and radius for
19 some other subclasses of starlike functions.
20
21
22
23 1. Introduction
24
25 Let An denote the class of functions f of the form
26
∞
27 X
28 f (z) = z + an+k z n+k ,
29 k=1
30
31 which are analytic in the open unit disk D = {z : |z| < 1, z ∈ C} . It is clear
32 that A1 = A is the class of normalized analytic functions. Also let S denote
33 the subclass of analytic functions A which are univalent in D. A function f
34 is said to be subordinate to a function g written as f ≺ g, if there exists a
35
36 Schwarz function w with w (0) = 0 and |w(z)| < 1 such that f (z) = g (w(z)) . In
37 particular, if g is univalent in D and f (0) = g (0), then f (D) ⊂ g ( D). Let ϕ
38 be analytic and univalent function in D such that ϕ (D) is convex with ϕ (0) = 1
39 and Re {ϕ0 (z)} > 0, z ∈ D. Then a function f ∈ A is in the class S ∗ (ϕ) , if it
40
41
satisfies
42 zf 0 (z)
≺ ϕ(z), z ∈ D.
43 f (z)
44
45 This class was introduced by Ma and Minda [21]. They also introduced the class
46 of convex functions as:
47 zf 00 (z)
 
48 C(ϕ) := f ∈ A : 1 + 0 ≺ ϕ(z) .
49 f (z)
50
51
For particular choices of function ϕ, we obtain several classes of analytic and
52 univalent functions. Janowski starlike function S ∗ [A, B] [20] is defined as
53  
∗ ∗ 1 + Az
54 S [A, B] := S , −1 ≤ B < A ≤ 1.
55 1 + Bz
56
57 Date: December 15, 2019.
58 ∗
Corresponding author
59 2010 Mathematics Subject Classification. Primary 30C45, 30C50.
60 Key words and phrases. Analytic functions, cosine functions, radii problems.
61
1
62
63
64
65
 1
2
3 2 K. BANO, M. RAZA
4
1+(1−2α)z
5 For ϕ (z) = 1−z
, 0 ≤ α < 1, the class of starlike function of order α defined
6 as
7
zf 0 (z)
 
8 ∗ 1 + (1 − 2α)z
9 S (α) := f ∈ A : ≺ , 0≤α<1 .
f (z) 1−z
10
11 Similarly the classes
12
13 
1 + Az

14 C[A, B] := C , −1 ≤ B < A ≤ 1,
15 1 + Bz
16
17 and
18  
19
1 + (1 − 2α)z
C(α) := C , 0≤α<1
20 1−z
21
22 are called the Janowski convex functions [20] and the class of convex functions
23 of order α, respectively. The classes S ∗ := S ∗ (0) and C := C(0) are well-known
24 1+z β


25 classes of starlike and convex functions respectively. For ϕ (z) = 1−z ,0<β≤
26 1, the class S ∗ (ϕ) reduces to the class SS ∗ (β) of strongly starlike function. The
27 class SP of parabolic starlike functions which was introduced by Rønning [2] is
28  √ 2
2 1+ z
29 obtained for the function ϕ (z) = 1 + π2 log 1−√z . For 0 ≤ k < ∞, the classes
30 k − ST of k-uniformly starlike functions and k − UCV of k-uniformly convex
31
32 functions was introduced by Kanas and Wisniowska [18] and can be obtained by
√
1 1+ z k2 −1
33 choosing ϕ (z) = 1−k 2 cosh A log 1− √
z
− 1−k 2 , where A = (2 cos k)/π. Sokól
34 √
35 and Stankiewicz [6] considered the class SL∗ := S ∗ ( 1 + z). The class SL∗ contains
0 (z)
36 the functions from A such that the quantity zff (z) maps open unit disk to the right
37 2
38 half plane of the lemniscate
√ of Bernoulli |ω − 1| < 1. Sokól [7] also introduced
39 the class Sq∗c = S ∗ ( 1 + cz), c ∈ (0, 1] which is related with the right loop of
40 the Cassinian ovals. Mendiratta et al. [14] introduced a subclass of starlike
41
functions associated
√ with left-half
q of the shifted lemniscate of Bernoulli, defined
42
∗
√ 
43 by SRL := S ∗ 2 − ( 2 − 1) 1+2(1−z √
2−1)z
. Sharma et al. [8] introduced the
44   √
2
45 class SC∗ := S ∗ 1 + 4z 3
+ 2z3 . For ϕ (z) = 1 + z + z, the class S ∗ (ϕ) reduces to
46
47 the class S(∗ of starlike function related with lune. The class S(∗ was introduced by
48 Raina and Sokól [16]. The class Se∗ = S ∗ (ez ) is studied by Mendiratta √ et al. [15].
49 ∗ ∗
Yunus et al. [22] introduced and studied the class Slim := S (1+ 2z +z 2 /2). For
50 z
ϕ (z) = 1 + 1−αz 2 , 0 < α ≤ 1, the class BS (α) of starlike functions related with
51
52 Booth lemniscate is introduced by Kargar et al. [17]. Goel and Kumar [12] studied
∗
53 the class SSG = S ∗ ( 1+e
2
z ) of starlike functions associated the sigmoid functions.
54 Wani and Swaminathan [9] introduced the class of starlike functions related with
55
56 nephroid functions. Cho et al. [11] considered the class Ss∗ to investigate the
57 geometric and radii problem for this class defined as
58
zf 0 (z)
 
59 ∗
60 Ss := f ∈ A : ≺ 1 + sin(z) .
61
f (z)
62
63
64
65
 1
2
3 COSINE FUNCTION 3
4
∗
5 Recently, Tang et al. [4] introduced the class Scos of starlike functions related
∗
6 with cosine functions. The class Scos of starlike functions is defined as:
7
8
zf 0 (z)
 
9 ∗
10 Scos := f ∈ A : ≺ cos(z) .
11
f (z)
12
13 ∗
The geometrical interpretation of the fact f ∈ Scos is that, for any z ∈ D, the
14 zf 0 (z)
15 ratio f (z) lies in the interior of the domain bounded by the disc with centre
cosh(1)+cos(1)
16
2
and radius cosh(1)−cos(1)
2
. Image of the cosine function under open unit
17
18 disc is given in the Figure 1.
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33 v 0
34 1
35 u

36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59 Figure 1. The geometrical representation of cos(z) with centre
cosh(1)+cos(1)
60 2
and radius cosh(1)−cos(1)
2
61
62
63
64
65
 1
2
3 4 K. BANO, M. RAZA
4
∗
5 It is clear that a function f is said to be in the class Scos if there exists an
6 anlytic function h, satisfying h(z) ≺ h0 (z) = cos(z), such that
7 Z z 
8 h(t) − 1
9
f (z) = z exp dt . (1.1)
0 t
10 ∗
11 Now we give some examples in the class Scos . For this, we consider the functions
12 z 4 + 2z 6 + 3z
13 h1 (z) = 1 + , h2 (z) = , h3 (z) = , h4 (z) = 1 + cos (1) z.
14 3 4+z 6+z
15 The function h0 (z) = cos(z) is univalent in D, hi (0) = h0 (0) = 1, (i = 1, 2, 3, 4)
16 and hi (D) ⊂ h0 (D). This implies that hi (z) ≺ h0 (z). Hence by using (1.1), we ob-
17 ∗
18 tain functions in the class Scos corresponding to every function hi (z) , (i = 1, 2, 3, 4)
19 respectively as follows:
20 z z2  z 2
21 f1 (z) = ze , f2 (z) = z + , f3 (z) = z 1 +
3 , f4 (z) = zecos(1)z .
22 4 6
23 If we take h (z) = h0 (z), then the function
24 Z z 
25 cos(t) − 1
26
f∗ (z) = z exp dt
0 t
27
28 plays the role of extremal function for many problems in the class S ∗cos .
29
∗
30 2. Starlikeness and Convexity of Functions in the Class Scos
31
32 In the following results, we investigate the starlikeness of order α and strongly
∗
33 starlikeness of order β for the class Scos . We also study the radius of convexity of
34 order β.
35
∗
36 Theorem 2.1. The class Scos satisfies the following relations:
37 ∗ ∗
(i) Scos ⊂ S (β), where β = cos(1).
38 ∗
⊂ SS ∗ (β), where β ≥ 2h eiθ0 /π ≈ 0.5060526391.


39 (ii) Scos
40 ∗
41
Proof. (i) Let f ∈ Scos . Then It is easy to see that
 0 
42 zf (z)
43 Re > minRe cos(z).
44
f (z) |z|=1

45 That is
46 minRe cos(z) = min cos(cos(θ)) cosh(sin(θ)).
47 |z|=1 −π≤t≤π
48 Let
49
50 ψ(t) = cos(cos(θ)) cosh(sin(θ)).
0
51 Then the equation ψ (θ) = 0 has five roots in the interval [−π, π] namely 0, ± π2
52 and ±π. Since ψ(θ) = ψ(−θ), so we consider the interval θ ∈ [0, π]. Furthermore,
53
54
we see that ψ(0) = cos(1) = ψ(π), ψ( π2 ) = cosh(1) and
55
h  π i
56 min ψ(0), ψ(π), ψ = ψ(π) = cos(1).
|z|=1 2
57
58 Thus, we have  0 
59 zf (z)
60 Re > cos(1).
61
f (z)
62
63
64
65
 1
2
3 COSINE FUNCTION 5
4
∗
5 (ii) Let f ∈ Scos . Then
6  0
  
7

arg zf (z) 
 < max arg cos (z) = max tan−1 − sin(cos(θ)) sinh(sin(θ))
.
8  f (z)  |z|=1 |z|=1 cos(cos(θ)) cosh(sin(θ))
9
10 Now
11 maxarg[cos(z)] = max h(eiθ ),
12 |z|=1 0≤θ≤2π
13 where  
14 iθ − sin(cos(θ)) sinh(sin(θ))
−1
15 h(e ) = tan .
16 cos(cos(θ)) cosh(sin(θ))
17 After simple calculations, it is observed that h0 (eiθ ) = 0 has two roots in [0, 2π],
18 namely
19
20 θ0 = 2.5291544179368 and θ1 = 0.6124382356911.
21 Thus the function h has maximum value at θ0 , as
22
23 h00 (θ0 ) < 0.
24

25 Hence max h(eiθ ) = h(eiθ0 ) = 0.5060526391. Thus
26  
∗ 2 iθ0


27 f ∈ SS h e .
28 π
29
30

31 ∗
Theorem 2.2. The C(β)-radius for the class f ∈ Scos is r0 , where r0 is the
32
33 smallest positive root of the equation
34 (cos(r) cosh(r) − β)(1 − r2 ) cosh(r) − r sinh(r) = 0,
35
36 where 0 ≤ β < 1.
37
∗
38 Proof. Since f ∈ Scos , there exists an analytic function w with w(0) = 0 and
39 |w(z)| ≤ |z| , such that
40
zf 0 (z)
41 = cos (w(z)) .
42 f (z)
43 After taking logarithmic differentiation, we get
44
45 zf 00 (z) zw0 (z) sin (w(z))
46 1+ = cos (w(z)) + .
f 0 (z) cos (w(z))
47
48 From the above relation, we have
49
zf 00 (z) |zw0 (z)| |sin (w(z))|
50 Re(1 + ) ≥ Re[cos (w(z))] − . (2.1)
51 f 0 (z) |cos (w(z))|
52
53 We assume that w(z) = Reiθ with R ≤ |z| = r, −π ≤ θ ≤ π for finding the mini-
54 mum of the right side of (4.1). First, we find the minimum value of Re(cos (w(z)))
55 with w(z) = Reiθ . A computation shows that
56
57 Re(cos(Reiθ )) = cos(Rx) cosh(Ry),
58
59 where x = cos(θ), y = sin(θ), x, y ∈ [−1, 1]. Now we observe that
60 cos(Rx) ≥ cos(R) ≥ cos(r) and 1 ≤ cosh(Ry) ≤ cosh(R) ≤ cosh(r).
61
62
63
64
65
 1
2
3 6 K. BANO, M. RAZA
4
5 Thus we conclude that
6
7 Re[cos w(z)] ≥ cos(r).
8
9 Now we consider that
10
sin(Reiθ )2 = cos(R cos(θ))2 sinh(R sin(θ))2 +sin(R cos(θ))2 cosh(R sin(θ))2 = φ(θ).
 
11
12
13 We observe that φ π2 = sinh2 (R) and φ(0) = sin2 (R) = φ(π). The equation


14
15 φ0 (θ) = 0 has five roots in [−π, π] namely 0, ± π2 and ±π. Since φ(θ) = φ(−θ),
16 therefore it is sufficient to consider θ ∈ [0, π]. Now
17 h  π i π 
18
19
max φ(0), φ(π), φ =φ = sinh2 (R).
2 2
20
21 Hence
22  
23 sin(Reiθ ) ≤ sinh(R) ≤ sinh(r).
24
25 Similarly, we can show that
26
27 cos(r) ≤ |cos (w(z))| ≤ cosh(r).
28
29 Thus we have
30
31
32 Re[cos (w(z))] ≥ cos(r), |sin(w(z))| ≤ sinh(r) and |cos (w(z))| ≤ cosh(r). (2.2)
33
34 Using all above inequalities and the following well-known inequality for Schwarz
35 function w [10], we obtain
36
37
0 1 − |w(z)|2 1 − R2 1
38 |w (z)| ≤ 2 = 2 ≤ .
39 1 − |z| 1 − |z| 1 − |z|2
40
41 From (2.2), we have
42
43 zf 00 (z) |zw0 (z)| |sin (w(z))|
 
r sinh(r)
44 Re 1 + 0 ≥ Re [cos (w(z))]− ≥ cos(r)− ≥ β.
45 f (z) |cos (w(z))| (1 − r2 ) cos(r)
46
47 After simplification, we get
48
49 (cos(r) − β)(1 − r2 ) cos(r) − r sinh(r) ≥ 0.
50
51 So the C(β)-radius r0 is the smallest positive root of the equation
52
53
54
(cos(r) − β)(1 − r2 ) cos(r) − r sinh(r) = 0,
55
56 where 0 ≤ β < 1. 
57
58 For β = 0, we have the following result.
59
60 ∗
Corollary 2.3. The radius of convexity for the class Scos is r0 ≈ 0.61938262.
61
62
63
64
65
 1
2
3 COSINE FUNCTION 7
4
5 3. Radius Problems
6
7 Here we discuss certain known classes to study the results in this section: Let
8 ( ∞
)
9 X 1 + Az
10 Pn [A, B] := p(z) = 1 + cn z n : p(z) ≺ , −1 ≤ B < A ≤ 1 .
k=n
1 + Bz
11
12 Also
13
14 Pn (α) := Pn [1 − 2α. − 1] and P n := Pn (0).
15 0

16 Now we take p(z) = zff (z)(z)

, for f ∈ An , then the class Pn [A, B] is denoted by
∗ ∗ ∗
17 Sn [A, B] and Sn := Sn [1 − 2α, −1]. Furthermore, let
18
∗ ∗
19 Scos,n := An ∩ Scos , Sn∗ (α) := An ∩ S ∗ (α), SL,n
∗
:= An ∩ SL∗ .
20
f (z)
21 Recently, Ali et al. [13] considered the classes, Sn := {f ∈ An : z
∈ Pn } and
22  
23 f (z) ∗
24 CS n (α) := f ∈ An : ∈ Pn , g ∈ Sn (α)
g(z)
25
∗
26 and obtained the SL,n -radii for the classes. They also obtained the conditions
27 on A and B so that Sn∗ [A, B] ⊂ SL,n∗
. Cho and Kumar obtained the starlikeness,
28 ∗
29 convexity and radii for different classes. In this section, Scos,n -radii for the classes
30 of janowski starlike functions and some other geometrically defined classes are
31 derived. We use the following results to establish our results.
32
33 Lemma 3.1. [3]If p ∈ Pn (α), then, for |z| = r,
34  0 
35  zp (z) 
≤ 2(1 − α)nrn
 p(z)  (1 − rn )(1 + (1 − 2α)rn ) .

36
37
38 Lemma 3.2. [19] If p ∈ Pn [A, B], then, for |z| = r,
39
2n  n
 
40
p(z) − 1 − ABr  ≤ (A − B)r .

41  1 − B 2 r2n  1 − B 2 r2n
42
43 In particular, if p ∈ Pn (α), then, for |z| = r,
44
2n  n
 
45 
p(z) − (1 + (1 − 2α))r  ≤ 2(1 − α)r .
46
47
 1 − r2n  1 − r2n
48 The purpose of our first lemma is to find the largest and the smallest disks
49
50
centered, respectively, at (a, 0) and (1, 0) such that the domain ∆c := C(D),
51 where C(z) = cos(z) is contained in the smallest disk and conatains the largest
52 disk.
53
54 Lemma 3.3. Let cos(1) < a < cosh(1). Then, the following inclusions hold:
55
56 {ω ∈ C : |ω − a| < ra } ⊂ ∆c ⊂ {ω ∈: |ω − 1| < cosh(1) − 1},
57
58 where
cos(1) < a ≤ cos(1)+cosh(1)

59 a − cos(1), 2
,
60 ra = cos(1)+cosh(1)
cosh(1) − a, 2
≤ a < cosh(1).
61
62
63
64
65
 1
2
3 8 K. BANO, M. RAZA
4
5 Proof. Let us first consider the square of the distance from (a, 0) to a point on
6 the boundary of ∆c , which is given by
7
8 h(t) = (a − cos(cos(θ)) cosh(sin(θ)))2 + (sin(cos(θ)) sinh(sin(θ)))2 .
9
10 In order to show that |ω − a|
p< ra is the largest disk contained in ∆c , we need
11 to show that only min0≤θ≤π h(θ) = ra . Since h(θ) = h(−θ), it is sufficient to
12 consider the range 0 ≤ θ ≤ π. We suppose that
13
14 cos(1) < a ≤ (1 + cos h(1) sin h(1))/(cos h(1) + sin h(1)).
15
16 It is easy to see that h0 (θ) = 0 has three roots 0, π2 and π. Also h0 (θ) > 0 for
17 0 ≤ θ ≤ π/2 and h0 (θ) < 0 for π/2 ≤ θ ≤ π. This implies
18 p p p
19 min0≤θ≤π h(θ) = h(0) = h(π) = a − cos(1).
20
21 We also suppose that
22 cos (1) + cosh(1)
23 (1 + cos h(1) sin h(1))/(cos h(1) + sin h(1)) < a ≤ .
24 2
Then h0 (θ) = 0 has five roots namely 0, θ0 ∈ 0, π2 , π2 , θ1 ∈ π2 , π and π. The



25
26 roots θ0 and θ1 depends upon a. The graph of h(θ) shows that
27 π

π

it is increasing
28 in the interval (0, θ0 ) , decresing in θ0 , 2 , increasing in 2 , θ1 and decreasing in
29 (θ1 , π) . Hence we conclude that
30 p p p
31 min0≤θ≤π h(θ) = h(0) = h(π) = a − cos(1).
32
Now we choose
33
34 cos (1) + cosh(1)
35
≤ a < 1 + cos(1) sin(1))/(cos(1) + sin(1).
2
36
Then h0 (θ) = 0 has five roots namely 0, θ2 ∈ 0, π2 , π2 , θ3 ∈ π2 , π and π. The



37
38 roots θ2 and θ3 depends upon a. The graph of h(θ) shows that

it is increasing
π π


39 in the interval (0, θ2 ) , decresing in θ2 , 2 , increasing in 2 , θ3 and decreasing in
40
41
(θ3 , π) . Hence we conclude that
r  
42 p π
43 min0≤θ≤π h(θ) = h = cosh(1) − a.
44 2
45 For the case
46
47 1 + cos(1) sin(1))/(cos(1) + sin(1) < a < cosh(1).
48
49 Then h0 (θ) = 0 has three roots namely π

0, 2 and π. The graph
of h(θ) shows that
π π
50 it is decreasing in the interval 0, 2 and increasing in 2 , π . Hence
51 r  
52 p π
53 min0≤θ≤π h(θ) = h = cosh(1) − a.
2
54
55 For circle of minimum radius centered at (1, 0) which contain f (D) = cos(z), we
56 find the maximum distance from (1, 0) to a point on the maximum distance from
57
(1, 0) to a point on the periphery of ∆c = f (D) and the square of this distance
58
59 function is given by
60 φ(θ) = (1 − cos(cos(θ)) cosh(sin(θ)))2 + (sin(cos(θ)) sinh(sin(θ)))2 .
61
62
63
64
65
 1
2
3 COSINE FUNCTION 9
4
5 The function φ attains its maximum value at point π2 , which is [cosh(1)−1]2 Thus
6 the radius of the smallest disk which contains ∆c = f (D) is cosh(1) − 1. 
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21 v 0
1
22 u
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48 Figure 2. The largest circle contained in ∆c = f (D)
49
50
51 ∗
Example 3.4. (a) The function f (z) = z + a2 z 2 is in Scos , if and only if
52
53 1 − cos(1)
54 |a2 | ≤ ≈ 0.3149.
55 2 − cos(1)
56 z ∗
57 (b) The function f (z) = 1−bz 2
is in Scos , if and only if
58
59 cosh(1) − 1
60 |b| ≤ ≈ 0.2136.
61
cosh(1) + 1
62
63
64
65
 1
2
3 10 K. BANO, M. RAZA
4
5
6
7
8
9
10
11
12
13
14
15
16
17 v 0
18 1
u
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42 Figure 3. The circle centered at (1, 0) with the smallest radius
43
44
cosh(1) containing ∆c = f (D)
45
46
47
48 Proof. (a) We know that f (z) = z + a2 z 2 ∈ S ∗ if and only if |a2 | ≤ 21 . Since
∗
49 Scos ⊂ S ∗ . we have |a2 | ≤ 21 , whenever f ∈ Scos
∗
. The function
50
51
52 zf 0 (z) 1 + 2a2 z
53
ω(z) = = ,
f (z) 1 + a2 z
54
55
56 maps D onto the disk
57
58 
2

59
 1 − 2|a |
2  |a2 |
ω − < .

60  1 − |a2 |  1 − |a2 |2
2
61
62
63
64
65
 1
2
3 COSINE FUNCTION 11
4
5 Now we are going to find conditions on a2 so that f (z) will be contain in ∆c .
2
6 2|
7
Since 1−2|a
1−|a |2
≤ 1, therefore
2
8
9 1 − 2|a2 |2 |a2 | 1 − 2 |a2 |2
10 cos(1) ≤ and ≤ − cos(1).
11
1 − |a2 |2 1 − |a2 |2 1 − |a2 |2
12 These two inequalities give
13
14
s
15 1 − cos(1) 1 − cos(1)
|a2 | ≤ and |a2 | ≤
16 2 − cos(1) 2 − cos(1)
17
18 respectively. Thus, we have
19
(s )
20
21
1 − cos(1) 1 − cos(1) 1 − cos(1)
|a2 | ≤ min , = .
22 2 − cos(1) 2 − cos(1) 2 − cos(1)
23
24 z
(b) Logarithmic differentiation of the function f (z) = 1−bz 2
yields
25
26 zf 0 (z) 1 + bz
27 ω(z) = = .
28 f (z) 1 − bz
29
30 We notice that the function ω maps D onto a disk
31 
 zf 0 (z) 1 + |b|2 

32 2 |b|
− ≤ .
 
33  f (z) 1 − |b|  1 − |b|2
2

34
35
It follows from previous Lemma that the above disk that will be contained in ∆c
36
37 provided
38
39 1 + |b|2 2 |b| 1 + |b|2
≤ cosh(1) and ≤ cosh(1) − .
40 1 − |b|2 1 − |b|2 1 − |b|2
41
42 These two inequalities give
43 s
44
cosh(1) − 1 cosh(1) − 1
45 |b| < and |b| ≤ ,
46 cosh(1) + 1 cosh(1) + 1
47
48 respectively. Thus, we have
49 (s )
50 cosh(1) − 1 cosh(1) − 1 cosh(1) − 1
51 |b| ≤ min , = .
52 cosh(1) + 1 cosh(1) + 1 cosh(1) + 1
53
54 This completes the proof. 
55
∗
56 Theorem 3.5. The sharp Scos,n -radius for the class Sn is given by
57
58 ! n1
59 1 − cos(1)
RScos,n
∗ (Sn ) = p .
60 n2 + (1 − cos(1))2 + n
61
62
63
64
65
 1
2
3 12 K. BANO, M. RAZA
4
5 Proof. Let f ∈ Sn . Consider the function h : D −→ C defined by
6
7 f (z)
h(z) = , h ∈ Pn .
8 z
9
10
Taking logarithmic differentiation, it follows that
11 zf 0 (z) zh0 (z)
12 −1= .
13 f (z) h(z)
14 Now an application of Lemma 3.1 gives
15  0   0 
 zf (z)   zh (z)  n
16
17
 − 1=  ≤ 2nr .
 f (z)   h(z)  1 − r2n
18
19 zf 0 (z)
In the view of Lemma 3.1, the image of |z| ≤ r under the function f (z)
will be
20
21 contained in the disk ∆c , if the following holds:
22 2nrn
23 ≤ 1 − cos(1).
24 1 − r2n
25 Solving the above inequality for r, we have
26
27 r2n (1 − cos(1)) + 2nrn + cos(1) − 1 ≤ 0.
28
∗
29 Thus, the Scos,n -radius of Sn is the smallest positive root of r2n (1−cos(1))+2nrn +
30 n)

31
cos(1) − 1 = 0 in (0, 1). Consider the function f0 (z) = z(1+z 1−z n
. Then it is easy
zf 0 (z) n
32 to see that Re f (z)
z
> 0 in the unit disk D. Thus, f0 ∈ Sn and f 0(z) = 1 + 1−z2nz
2n .
33
34
Also f0 ensures the sharpness of the result since at z = RScos,n
∗ (Sn ), we have
35
zf00 (z) 2nz n
36 −1= = 1 − cos(1).
37 f0 (z) 1 − z 2n
38
39 This completes the proof. 
40 ∗
41
Theorem 3.6. The Sharp Scos,n -radius for the class CS n (α) is given by
42 ! n1
43 1 − cos(1)
44 RScos,n
∗ (CS n (α)) = p .
45 (n + 1 − α) + (n + 1 − α)2 + (1 − cos(1)) (1 + cos(1) − 2α)
46 f (z)
47 Proof. Let f ∈ CS n (α). Then define h(z) = g(z)
, where g ∈ S ∗ (α) and h ∈ Pn .
48 From the definition of h, we have
49
50 zf 0 (z) zg 0 (z) zh0 (z)
51 = + .
f (z) g(z) h(z)
52
53 Using Lemma 3.2, we have
54  0
 zg (z) 1 + (1 − 2α)r2n  2(1 − α)rn

55
 g(z) −
 ≤ .
56 1 − r2n  1 − r2n
57
58 Therefore  0
 zf (z) 1 + (1 − 2α)r2n  2(1 + n − α)rn

59
 f (z) −
60
 ≤ . (3.1)
61
1 − r2n  1 − r2n
62
63
64
65
 1
2
3 COSINE FUNCTION 13
4
2n
5 Now centre of the above disc 1+(1−2α)r
1−r2n
≤ cos(1)+cosh(1)
2
if and only if r2n ≤
6 cos(1)+cosh(1)−2
7 2−4α+cos(1)+cosh(1)
0 ≤ α < 1. Let r ≤ RScos,n
∗ ( CS n (α)). Then
8 !2
9 1 − cos(1)
10 r2n ≤ p
11 (2 − α) + (2 − α)2 + (1 − cos(1)) (1 + cos(1) − 2α)
12 cos (1) + cosh(1) − 2
13 ≤ .
14 2 − 4α + cos (1) + cosh(1)
15
16 Therefore the centre of (3.1) is less than cos(1)+cosh(1)
2
. It follows from Lemma 3.3
17 that
18 2(1 + n − α)rn 1 + (1 − 2α)r2n
19 ≤ − cos(1).
1 − r2n 1 − r2n
20
21
Which is equivalent to
22 r2n (1 + cos(1) − 2α) − 2(1 + n − α)rn + 1 − cos(1) ≤ 0.
23
24 Since r ≤ RScos,n
∗ (CS n (α)), therefore we have the reuired result. Consider the
25 functions f∗ and g∗ given as
26
27 z (1 + z n ) z
f∗ (z) = (n+2−2α)/n
and g∗ (z) = .
28 (1 − z )n (1 − z )(2−2α)/n
n
29
30 Now
31 
f∗ (z)
 
1 + zn
  0
zg∗ (z)
 
1 + (1 − 2α)z 2n

32 Re = Re > 0 and Re = Re > α.
33 g∗ (z) 1 − zn g∗ (z) 1 − z 2n
34 This shows that the function f∗ satisfies the hypothesis of Theorem 3.6. If we
35
36 choose z = RScos,n
∗ (CS n (α)) for every n ∈ N, then
37
zf∗0 (z) 1 + 2 (1 + n − α) z n + (1 − 2α) z 2n
38 = = 1 − cos(1).
39 f∗ (z) 1 − z 2n
40
41
This shows that result is sharp. 
42 ∗
Theorem 3.7. Let 0 ≤ B < A ≤ 1. Then Scos,n radius for the class Sn∗ [A, B] is
43
44 given by
  1/n 
45 1 − cos(1)
46 RScos,n
∗ = min 1; .
A − B cos(1)
47
48 Proof. Let f ∈ Sn∗ [A, B]. Then in the view of Lemma 3.2, we have
49  0 n

 ≤ A − Br ,
50  zf (z) 
51

 f (z) − b  1 − B 2 r2n (3.2)
52
53 where the center of the disk is b = (1 − ABr2n ) / (1 − B 2 r2n ). We note that b ≤ 1
54 for B ≥ 0. By using Lemma 3.3, we have
55
56 A − Brn 1 − ABr2n
57 ≤ − cos(1),
1 − B 2 r2n 1 − B 2 r2n
58
59 which is equivalent to
60 B(A − B cos(1))r2n + (A − B)rn − 1 + cosh(1) ≤ 0.
61
62
63
64
65
 1
2
3 14 K. BANO, M. RAZA
4
∗
5 Solving the above inequality, we get r ≤ RScos
∗ (S [A, B]). Furthermore b = 1 for
n
6 B = 0 and from equation (3.2), we have
7  0 
8  zf (z)  n
 f (z) − 1 ≤ Ar 0 < A ≤ 1.
 
9
10
1
11 Using Lemma 3.3, when a = 1 leads to f ∈ Scos,n ∗
, if r ≤ ( 1−cos(1)
A
) n . Result is
12
sharp for the function
13  A−B
14 z (1 + Bz n nB
) , B 6= 0,
15 f0 (z) =
16 z exp (Az n /n) , B = 0.
17 Then it is claer that zf00 (z)/f0 (z) = (1 + Az n ) / (1 + Bz n ). Hence f0 (z) ∈
18
19 Sn∗ [A, B]. 
20
21
Theorem 3.8. Let −1 < B < A ≤ 1. Then Sn∗ [A, B] ⊂ Scos,n ∗
if the parameters
22 A and B satisfy either of the relation
23 (a) 2(1 − B 2 ) cos(1) < 2 (1 − AB) ≤ (cos(1) + cosh(1)) (1 − B 2 ) and (1 −
24 B) cos(1) ≤ (1 − A), (b) (cos(1) + cosh(1)) (1 − B 2 ) < 2 (1 − AB) < 2 cosh(1)(1 −
25
B 2 ) and (1 + A) ≤ (1 + B) cosh(1).
26
27 0

28
Proof. Let f ∈ Sn∗ [A, B]. Then zff (z)
(z)
∈ Pn [A, B], in view of Lemma 3.2, we have
29  
p(z) − 1 − AB  ≤ A − B .
 
30
31
 1 − B  1 − B2
2

32
33 The above inequality represents a disc with centre (1 − AB) / (1 − B 2 ) and radius
34 (A − B) / (1 − B 2 ) . To prove our result we have to show that
35 (
1−AB
cos(1) < 1−AB ≤ cos(1)+cosh(1)
 
36 
p(z) − 1 − AB 
 ≤ ra = 1−B 2
− cos(1), 1−B 2 2
,
37 2 1−AB cos(1)+cosh(1) 1−AB
 1−B  cosh(1) − 1−B 2 , 2
≤ 1−B 2 < cosh(1).
38
39 This implies that
40 (
41 A−B 1−AB
1−B 2
− cos(1), cos(1) < 1−AB
1−B 2
≤ cos(1)+cosh(1)
2
,
42 ≤ ra = cos(1)+cosh(1)
43
1 − B2 cosh(1) − 1−AB
1−B 2
, 2
1−AB
≤ 1−B 2 < cosh(1).
44 Which is equivalent to either
45
46 A−B 1 − AB 1 − AB cos (1) + cosh(1)
47 2
≤ 2
− cos(1) and cos(1) < 2
≤
1−B 1−B 1−B 2
48
49 or
50 A−B 1 − AB cos (1) + cosh(1) 1 − AB
51 ≤ cosh(1) − and ≤ < cosh(1).
1 − B2 1 − B2 2 1 − B2
52
53 After some simplifications, we have the required result. 
54 ∗
55 Theorem 3.9. The sharp Scos -radii for the classes SL∗ , SRL
∗
, SC∗ , Se∗ and Slim
∗
are:
∗ 2
∗ (S ) = sin (1) ≈ 0.70807,
56 (1) RScos L √ √ √
57 ∗ 4 2+5−2(3 2+4) cos(1)+(3+2 2)cos(1)2
(2) RScos
∗ (S
RL ) = 5+4 2−4( 2+2) cos(1)+2(1+ 2)cos(1)2 ≈ 0.73626,
√ √ √
58
√
 
59 ∗ 1 1 −1
(3) RScos
∗ (S ) = 2
− 2e 2 + −2e + 3e + 3 e 2 ≈ 0.347078,
60 C 2
61
62
63
64
65
 1
2
3 COSINE FUNCTION 15
4
∗
∗ (S ) = ln(cosh(1)) ≈ 0.43378,
5 (4) RScos e
6 ∗
√ q
e2 +1
7 (5) RScos
∗ (S
lim ) = − 2 + e
≈ 0.3426.
8 0 (z) √
9 Proof. (1) Let f ∈ SL∗ . Then, we have zff (z) ≺ 1 + z. Thus for |z| ≤ r <
10 ∗
11 RScos
∗ (S ), we have
L
12  √ √
 0  
 zf (z) 
13 − 1 = 1 + z − 1 ≤ 1 − 1 − r ≤ 1 − cos(1).
  
 f (z)   
14
15 By using Lemma 3.3, we have the required result. To show the sharpness of the
16
17 result, we consider the function defined by
18
 √ 
4z exp 2( 1 + z − 1)
19 f0 (z) = √ .
20 (1 + 1 + z)2
21 zf 0 (z) √ zf 0 (z) √
22 Since f00(z) = 1 + z, it follows that f0 ∈ SL∗ . Also f00(z) = 1 + z = cos(1) at
23
z = sin2 (1) and hence sharpness of the result is verified. ( see Fig.4).
24
25
∗
26 (2) For the function f ∈ SRL , we have
27 s
28 zf 0 (z) √ √ 1−z
29 ≺ 2 − ( 2 − 1) √ ,
f (z) 1 + 2( 2 − 1)z
30
31 ∗
Thus for |z| ≤ r < RScos
∗ (S
RL ), we have
32  
33
s
 0
 zf (z)

 √ √ 1 − z 
34
 f (z) − 1 =  2 − ( 2 − 1) 1 + 2(√2 − 1)z − 1
   
35
36 s !
37 √ √ 1+r
38 ≤ 1− 2 − ( 2 − 1) √ ≤ 1 − cos(1).
39 1 − 2( 2 − 1)r
40
41 By using Lemma 3.3, we obtain the required result. To show the sharpness of the
42 result, we consider the following function defined as
43 Z z 
q0 (t) − 1
44 f0 (z) = z exp dt ,
45 0 1
46
47
where s
48 √ √ 1−z
49 q0 (z) = 2 − ( 2 − 1) √ .
1 + 2( 2 − 1)z
50
51 zf 0 (z) ∗ zf 0 (z) √ √
Since f00(z) = q0 (z), it follows that f0 ∈ SRL . Also f00(z) = 2−( 2−
52 q
53 1) 1+2(1−z
√
2−1)z
= 1 − cos(1) at
54
55 √ √ √
56 4 2 + 5 − 2(3 2 + 4) cos(1) + (3 + 2 2)cos(1)2
z= √ √ √
57 5 + 4 2 − 4( 2 + 2) cos(1) + 2(1 + 2)cos(1)2
58
59 and hence sharpness of the result is verified.( see Fig.5)
0 (z) 2
60 (3) Suppose f ∈ SC∗ , then zff (z) ≺ 1 + 4z
3
+ 2z3 , Thus for |z| = r, gives
61
62
63
64
65
 1
2
3 16 K. BANO, M. RAZA
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18 v 0
1
19 u
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42 ∗
43 Figure 4. Scos radii for the class SL∗
44
45
46  0   2

47  zf (z)   4z 2z 
 f (z) − 1 = 1 + 3 + 3 − 1
   
48
49
4r 2r2
 
50
51
≤ 1+ + − 1 ≤ cosh(1) − 1
3 3
52
53
√
 
54 1 1 −1

55 for r ≤ 2
− 2e + 2 −2e + 3e2 + 3 e 2 . To verify the sharpness of the result,
56 consider the function f0 defined by
57
58
4z + z 2
 
59
60 f0 (z) = zexp .
61
3
62
63
64
65
 1
2
3 COSINE FUNCTION 17
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18 v 0
1
19 u
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42 ∗ ∗
43 Figure 5. Scos radii for the class SRL
44
45
46 zf00 (z) 2

47 Since f0 (z)
= 1 + 4z
3
+ 2z3 , it follows that f0 ∈ SC∗ and at z = RScos ∗
∗ (S ), we have
C
48
49 zf00 (z)
50 = cosh(1).
51 f0 (z)
52
53 Hence the result is sharp.(see Fig.6)
54 (4) For f ∈ Se∗ , we have
55
56  0
 zf (z)


57 z r
 f (z) − 1 = |e − 1| ≤ e − 1 ≤ cosh(1) − 1.
 
58
59
60 The result is sharp for the function.( see Fig.7)
61
62
63
64
65
 1
2
3 18 K. BANO, M. RAZA
4
5
6
7
8
9
10
11
12
13
14
15
16 v 0
1
17 u
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42 ∗
43 Figure 6. Scos radii for the class SC∗
44
45 0 (z) √ 2
46 (5)Suppose f ∈ SC∗ , then zff (z) ≺ 1 + 2z + z2 , Thus for |z| = r, gives
47
√
48
 0   2

 zf (z)   z 
49 
 f (z) − 1 = 1 + 2z + − 1
50
  2 
51
√
2r r2
 
52 ≤ 1+ + − 1 ≤ cosh(1) − 1
53 3 2
54
55 √ q e2 +1
for r ≤ − 2 + e
. To verify the sharpness of the result, consider the function
56
57 f0 defined by
58 ( √ )
59 4 2z + z 2
f0 (z) = zexp .
60 4
61
62
63
64
65
 1
2
3 COSINE FUNCTION 19
4
5
6
7
8
9
10
11
12
13
14
15
16 v 0
1
17 u
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42 ∗
43 Figure 7. Scos radii for the class Se∗
44
45 zf00 (z) √ z2 ∗ ∗
46 Since f0 (z)
=1+ 2z + 2
, it follows that f0 ∈ Slim and at z = RScos
∗ (S
lim ), we
47 have
48 zf00 (z)
49 = cosh(1).
50 f0 (z)
51
52
Hence the result is sharp.(see Fig.8) 
53
54 Consider
n the functions    o
55 F1 = f ∈ An : Re fg(z) (z)
> 0 and Re g(z) > 0, g ∈ A n ,
56 n    z  o
57 F2 = f ∈ An : Re fg(z) (z)
> 0 and Re g(z) > 1/2, g ∈ A n ,
58 n    z o
F3 = f ∈ An :  fg(z)
 (z)
− 1 < 1 and Re g(z) > 0, g ∈ An .

59 z
60
61
62
63
64
65
 1
2
3 20 K. BANO, M. RAZA
4
5
6
7
8
9
10
11
12
13
14
15
16 v 0
1
17 u
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42 ∗ ∗
43 Figure 8. Scos radii for the class Slim
44
45
46 ∗
Theorem 3.10. The sharp Scos,n -radii for the functions in the classes F1 , F2 , F3 and F4
47
48
respectively are
  n1
49 1−cos(1)
50 (1) RScos,n
∗ (F1 ) = √ 2 2
,
2n+ 4n +(1−cos(1))
51   n1
52 2(1−cos(1))
53 (2)RScos,n
∗ (F2 ) = √ ,
9n2 +4n−4n cos(1)−8 cos(1)+4cos(1)2 +4+3n
54   n1
55 2(1−cos(1))
56 (3)RScos,n
∗ (F3 ) = √ .
9n2 +4n−4n cos(1)−8 cos(1)+4cos(1)2 +4+3n
57
58
59 Proof. (1) Let f ∈ F1 and define the functions p, h : D → C by p(z) = g(z)z
and
f (z)
60 h(z) = g(z) . Then, clearly p, h ∈ Pn . Since f (z) = zp(z)h(z), it follows from
61
62
63
64
65
 1
2
3 COSINE FUNCTION 21
4
5 Lemma 3.1 that  0 n

 zf (z)
6
 ≤ 4nr ≤ 1 − cos(1)

7

 f (z) − 1 1 − r2n
8
9   n1
10 for r ≤ √ 1−cos(1)
2 2
= RScos,n
∗ (F1 ), to verify the sharpness of the result,
11 2n+ 4n +(1−cos(1))
12 consider the functions
13 2
1 + zn 1 + zn
  
14 f0 (z) = z and g0 (z) = z .
15 1 − zn 1 − zn
16
17 Thus, clearly    
18 f0 (z) g0 (z)
19 Re > 0 and Re >0
g0 (z) z
20
iπ
21 and hence f ∈ F1 . A computation shows that at z = RScos,n
∗ (F1 )e n
22
23 zf 0 (z) 4nz n
24 =1+ = 2 − cos(1).
25
f (z) 1 − z 2n
26 This confirms the sharpness of the result.
27
28
(2) Let f ∈ F2 . Define the function p, h : D → C by p(z) = g(z) z
and h(z) = fg(z)
(z)

29 . Then, p ∈ Pn and h ∈ Pn ( 12 ). Since f (z) = zp(z)h(z), it follows from Lemma

30 3.1 that  0
 nrn (3 + rn )

31  zf (z)
32  − 1= ≤ 1 − cos(1).
33
 f (z)  1 − r2n
34 Hence
35   n1
36
37 2(1 − cos(1))
r ≤ q  = RS ∗ (F2 )
cos,n
38 2 2
39 9n + 4n − 4n cos(1) − 8 cos(1) + 4cos(1) + 4 + 3n
40 iπ
∗
41 Thus f ∈ Scos,n for r ≤ RScos,n
∗ (F2 )e n .
42 To see the sharpness of the result, consider the functions
43
44 z(1 + z n ) z
f0 (z) = n 2
and g0 (z) = .
45 (1 − z ) 1 − zn
46
47 Then clearly Re( f0g(z)
0
) > 0 and Re( g0z(z) ) > 12 and hence f ∈ F2 . The Sharpness
48
49 is evident, since at z = RScos,n
∗ (F2 ), we have
50
zf00 (z) nz n (3 + z n )
51 −1= = 1 − cos(1).
52 f0 (z) 1 − z 2n
53
54 ( 3) Let f ∈ F3 .Taking functions p, h defined
 on open
 unit disk by p(z) = g(z)
z
and
55
h(z) = fg(z) 1
. Then p ∈ Pn . We note that  h(z) − 1 < 1 if and only if Re(h(z)) > 21
 
56 (z)
57 and therefore h ∈ Pn ( 12 ). Using Lemma 3.1, we have
58  0
 3nrn + nr2n

59  zf (z)
60
 − 1≤ .
61
 f (z)  1 − r2n
62
63
64
65
 1
2
3 22 K. BANO, M. RAZA
4
5 By using Lemma 3.3, we have
6
7 3nrn + nr2n
≤ 1 − cos(1).
8 1 − r2n
9
10 For sharpness consider the functions
11
12
z(1 + z n )2 z(1 + z n )
f0 (z) = and g0 (z) = .
13 1 − zn 1 − zn
14 Since
15 
 f0 (z)
  
16
 n g0 (z)
 g0 (z) − 1 = |z | < 1 and Re > 0.
 
17 z
18 iπ
19 Hence f0 ∈ F3 . Now at z = RSc,n
∗ (F3 )e n , we have

20
21 zf00 (z) nz 2n − 3nz n
1− = = 1 − cos(1).
22 f0 (z) 1 − z 2n
23
24 This confirms the sharpness of the result.
25 
26
∗
27 Theorem 3.11. Let f ∈ Scos , then f ∈ S ∗ (cos(1)) ∩ M(cosh(1)).
28
29 ∗ zf 0 (z)
Proof. Let f ∈ Scos , we have f (z)
≺ cos(z). It is easy to note that
30
31 zf 0 (z)
32 minRe cos(z) < Re < maxRe cosh(z).
33 |z|=1 f (z) |z|=1
34 This implies that
35
36 zf 0 (z)
cos(1) < Re < cosh(1).
37 f (z)
38
39 Thus
40 f ∈ S ∗ (cos(1)) ∩ M(cosh(1)).
41
42 
43
44 4. Differential Subordination Results
45
46 Lemma 4.1. [5] Let w be a non constant analytic function in D with w (0) = 0.
47
If
48
49 |w (z0 )| = max {|w(z)| , |z| ≤ |z0 |} , z ∈ D,
50
51
then there exists a real number m (m ≥ 1) such that z0 w0 (z0 ) = mw (z0 ) .
52 (A−B)(cosh(1)) k
53 Theorem 4.2. Assume that |β| ≥ sin(1)−|B| sinh(1)
and p is an analytic function
54 defined on D with p(0) = 1 satisfying
55
56 zp0 (z) 1 + Az
57 1+β ≺ , −1 ≤ B < A ≤ 1, k = 0, 1, 2 · · · .
pk (z) 1 + Bz
58
59 In addition suppose that p(z) = cos (w (z)), where w is an analytic function in D
60 with w(0) = 0. Then p(z) ≺ cos (z).
61
62
63
64
65
 1
2
3 COSINE FUNCTION 23
4
5 Proof. Consider the functions
6
7 zp0 (z)
h(z) = 1 + β , z∈D (4.1)
8 pk (z)
9
10 and
11
p(z) = cos (w (z)) , (4.2)
12
13 where w is analytic function in D with w (0) = 0. It suffices to show that |w(z)| <
14 1. From (4.1) and (4.2) , we have
15
16 βzw0 (z) sin (w (z))
17 h(z) = 1 − .
18 [cos (w (z))]k
19
20 This implies that
21 h(z) − 1 βzw0 (z) sin (w (z))
22 = .
23 A − Bh(z) (A − B) [cos (w (z))]k + βBzw0 (z) sin (w (z))
24
25 We suppose that there exists a point z0 ∈ D such that max |w(z)| = |w (z0 )| = 1.
|z|≤|z0 |
26
27 By using Lemma 4.1, there exists a number m ≥ 1 such that z0 w0 (z0 ) = mw (z0 ).
28 We also suppose that w (z0 ) = eiθ . Then we have
29  
0
 
30  h(z) − 1   −βzw (z) sin (w (z)) 
31
  =   
 A − Bh(z)  k 
 (A − B) [cos (w (z))] + βBzw0 (z) sin (w (z)) 
32
33 |β| m |sin (w (z))|
34 ≥ .
35 (A − B) |cos (eiθ )|k + |β| |B| m |sin (eiθ )|
36
37 Let |z| = r, −π ≤ θ ≤ π. A computation shows that
38
sin eiθ 2 = cos2 (cos θ) sinh2 (sin θ) + sin2 (cos θ) cosh2 (sin θ) = Ψ (θ) . (4.3)


39
40
41 The equation Ψ0 (θ) = 0 has five roots in [−π, π] namely 0, ±π, ± π2 . Since Ψ0 (θ) =
42 Ψ0 (−θ) , it is sufficient to consider θ ∈ [0, π] and we see that
43 π 
44 Ψ00 (0) = Ψ00 (π) ≈ 1.090702574 and Ψ00 ≈ −1.626860410.
45 2
46
47
This implies that
48 π 
49 max {Ψ (θ)} = Ψ = sinh2 (1),
50
2
51 min {Ψ (θ)} = Ψ (0) = Ψ (π) = sin2 (1).
52
53 Also consider
54 cos eiθ 2 = cos2 (cos θ) cosh2 (sin θ) + sin2 (cos θ) sinh2 (sin θ) = φ (θ) .


55
56
57
The equation φ0 (θ) = 0 has five roots in [−π, π] namely 0, ±π, ± π2 . Since φ (θ) =
58 φ (−θ) , it is sufficient to consider those roots which lies in [0, π] and we see that
59 π 
60 φ00 (0) = φ (π) ≈ 2.909297427 and φ00 ≈ −5.626860410
61 2
62
63
64
65
 1
2
3 24 K. BANO, M. RAZA
4
5 and
6 π 
7 max {φ (θ)} = φ = cosh2 (1),
8
2
9 min {φ (θ)} = φ (π) = φ (0) = cos2 (1).
10
11 Therefore
 
12 cos (1) ≤ cos eiθ  ≤ cosh (1) , sin(1) ≤ sin eiθ  ≤ sinh(1).



(4.4)
13
14 This implies that
15  
16  h(z) − 1  |β| m sin(1)
  ≥
17  A − Bh(z)  (A − B) [cosh (1)]k + |β| |B| m sinh(1)
18
19 = F (m) (say).
20
21 Now
22 |β| sin(1) (A − B) [cosh (1)]k
23
F 0 (m) = h i2 .
k
24 (A − B) [cosh (1)] + |β| |B| m sinh(1)
25
26 Then the function F 0 (m) > 0. This implies that F (m) ≥ F (1) . Therefore
27
 
 h(z) − 1  |β| sin(1)
28  A − Bh(z)  ≥ (A − B) [cosh (1)]k + |β| |B| sinh(1) .
 
29
30
31 The right hand side of the above inequality is greater than 1, when |β| ≥
(A−B)(cosh(1))k
32
sin(1)−|B| sinh(1)
. This contradicts our supposition. Hence p(z) ≺ cos(z). This
33
34
completes the proof. 
35 For A = 1, B = 0, we have the following result.
36
k
37
38
Corollary 4.3. Assume that |β| ≥ (cosh(1))
sin(1)
and p is an analytic function defined
39 on D with p(0) = 1 satisfying
40
 zp0 (z) 
 
41
 pk (z)  < 1, k = 0, 1, 2 · · · .
β 
42
43
44 In addition suppose that p(z) = cos (w (z)), where w is an analytic function in D
45 with w(0) = 0. Then p(z) ≺ cos (z).
46 k
(A−B)(cosh(1))
47 Corollary 4.4. Assume that |β| ≥ sin(1)−|B| sinh(1)
and f is an analytic function
48
49 defined on D with p(0) = 1 satisfying
50 zp0 (z) 1 + Az
51 1+β k
≺ , −1 ≤ B < A ≤ 1, k = 0, 1, 2 · · · .
52 p (z) 1 + Bz
53 In addition suppose that p(z) = cos (w (z)), where w is an analytic function in D
54
55 with w(0) = 0. Then p(z) ≺ cos (z).
56 (A−B)(cosh(1)) k
57 Corollary 4.5. Let |β| ≥ sin(1)−|B| sinh(1)
and f ∈ A. Assume that
58
(zf 0 (z))0 zf 0 (z)
   0 1−k
59 zf (z) 1 + Az
60 1+β 0
− ≺ .
61
f (z) f (z) f (z) 1 + Bz
62
63
64
65
 1
2
3 COSINE FUNCTION 25
4
5 In addition suppose that
6 zf 0 (z)
7 = cos (w (z)) ,
8 f (z)
∗
9 where w is an analytic in D with w(0) = 0. Then f ∈ Scos .
10
0
11 Proof. Consider p(z) = zff (z)
(z)
, then it is clear that p is analytic in D and p(0) =
12
13 1. After simple calculations, we have
(zf 0 (z))0 zf 0 (z)
14   0 k−1
zp0 (z)

zf (z)
15 1+β k =1+β 0
− .
16 p (z) f (z) f (z) f (z)
17
18
Since, it is given that
(zf 0 (z))0 zf 0 (z)
19    0 k−1
zf (z) 1 + Az
20 1+β 0
− ≺ ,
21 f (z) f (z) f (z) 1 + Bz
22 ∗
23
therefore by using Theorem 4.2, f ∈ Scos . 
24
25 References
26
27 [1] Altinkaya S. Yalçin S. Third Hankel determinant for Bazilevi č functions. Advances in
28 Mathematics 2016; 5(2): 91-96.
29 [2] F. Rønning, Uniformly convex functions and a corresponding class of starlike functions,
30 Proc. Amer. Math. Soc. 118 (1993), no. 1, 189–196.
31 [3] G. M. Shah, On the univalence of some analytic functions. Pacific J. Math., 43(1)(1972),
32 239-250.
33 [4] H. Tang, H. M. Srivastava, S-H. Li and G-T. Deng, Majorization results for
34 subclasses of starlike functions based on the sine and cosine functions, (2019).
35 https://doi.org/10.1007/s41980-019-00262-y.
36 [5] I. S. Jack, Functions starlike and convex of order α, J. London Math. Soc., 3(1971), 469-474.
37 [6] J. Sokól and J. Stankiewicz, Radius of convexity of some subclasses of strongly starlike
38 functions, Zeszyty Nauk. Politech. Rzeszowskiej Mat, 19(1996), 101-105.
39 [7] J. Sokól, On some subclass of strongly starlike functions, Demonstratio Math. 31(1998),
40 81–86.
41 [8] K. Sharma N. K. Jain and V. Ravichandran, Starlike functions associated with a cardioid,
42
Afr. Math., 27(5-6)(2016), 923-939.
43
[9] L. A. Wani and A. Swaminathan, Starlike and convex functions associated with a nephroid
44
45 domain having cusps on the real axis, arXiv:1912.05767.
46 [10] Nehari, Z. (2012). Conformal mapping. Courier Corporation.
47 [11] N. E. Cho, V. Kumar, S. S. Kumar and V. Ravichandran, Radius problems for starlike
48 functions associated with the sine function, Bull. Iran Math. Soc., 45(1)(2019), 213-232.
49 [12] P. Goel and S. S. Kumar, Certain class of starlike functions associated with modified
50 sigmoid function, Bull. Malays. Math. Sci. Soc. (2019). https://doi.org/10.1007/s40840-
51 019-00784-y.
52 [13] R. M. Ali, N. K. Jain and V. Ravichandran, Radii of starlikeness associated with the
53 lemniscate of Bernoulli and the left-half plane, Appl. Math. Comp., 218(11)(2012), 6557-
54 6565.
55 [14] R. Mendiratta, S. Nagpal and V. Ravichandran, A subclass of starlike functions associated
56 with left-half of the lemniscate of Bernoulli, Int. J. Math., 25(09)(2014), 1450090.
57 [15] R. Mendiratta, S. Nagpal and V. Ravichandran, On a subclass of strongly starlike functions
58 associated with exponential function, Bull. Malays. Math. Sci. Soc. 38 (2015), 365–386.
59 [16] R. K. Raina and J. Sokól, Some properties related to a certain class of starlike functions,
60 C. R. Math. Acad. Sci. Paris 353 (2015), no. 11, 973–978.
61
62
63
64
65
 1
2
3 26 K. BANO, M. RAZA
4
5 [17] R. Kargar, A. Ebadian and J. Sokól, On Booth lemniscate and starlike functions, Anal.
6 Math. Phys. 9 (2019), 143–154.
7 [18] S. Kanas and A. Wisniowska, Conic regions and k-uniform convexity, J. Comput. Appl.
8 Math. 105(1999), 327–336.
9 [19] V. Ravichandran, F. Ronning and T. N. Shanmugam, Radius of convexity and radius of
10 starlikeness for some classes of analytic functions, Comp. Var. Ellip. Equ., 33(1-4)(1997),
11 265-280.
12 [20] W. Janowski, Extremal problems for a family of functions with positive real part and for
13 some related families, Ann. Polon. Math., 23(1970), 159-177.
14
[21] W. Ma and D. Minda, A unified treatment of some special classes of univalent functions. In
15
Proceeding of the Conference on Complex Analysis, Z. Li, F. Ren, L. Yang and S. Zhang
16
17 (Eds), Int. Press (1994) (pp. 157-169).
18 [22] Y. Yunus, S. A. Halim and A. B. Akbarally, Subclass of starlike functions associated with
19 a limacon, in AIP Conference Proceedings 2018 Jun 28 (Vol. 1974, No. 1, p. 030023), AIP
20 Publishing.
21 1
22 Department of Mathematics, Government College University Faisalabad,
23 Pakistan
24 Email address: [email protected]
25 2
26 Department of Mathematics, Government College University Faisalabad,
27 Pakistan
28 Email address: [email protected]
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65