Introduction
In the year 2008, Aytar [2 , 3 ] introduced the notion of rough statistical convergence as a generalization of rough convergence [28 , 29 ] and statistical convergence [32 , 35 ] . He demonstrated rough statistical convergence in the following manner: A sequence x = { x n } n ∈ N x = \{x_n\}_{n\in\mathbb{N}} x = { x n } n ∈ N in a norm linear space X X X is said to be rough statistically convergent to x ∗ x_* x ∗ w.r.t the roughness of degree r r r (where r ≥ 0 r \ge0 r ≥ 0 ) if for every ε > 0 \varepsilon> 0 ε > 0 ,
lim n → ∞ 1 n ∣ { k ≤ n : ∥ x k − x ∗ ∥ ≥ r + ε } ∣ = 0 , (1) \lim_{n\to\infty} \frac{1}{n}\left|\{k\le n : \lVert x_k-x_*\rVert\ge r+\varepsilon\}\right|=0, \tag*{(1)} n → ∞ lim n 1 ∣ { k ≤ n : ∥ x k − x ∗ ∥ ≥ r + ε } ∣ = 0 , (1)
and we write x n → s t r x ∗ x_n \xrightarrow{st}_{r} x_* x n s t r x ∗ . Thus, the set s t − LIM r x = { x ∗ ∈ X : x n → s t r x ∗ } st\!-\!\operatorname{LIM}^{r}x=\{x_*\in X:x_n\xrightarrow{st}_{r}x_*\} s t − LIM r x = { x ∗ ∈ X : x n s t r x ∗ } is called the rough statistical limit set of the sequence x = { x n } n ∈ N x=\{x_n\}_{n\in\mathbb{N}} x = { x n } n ∈ N with degree of roughness r r r . Moreover Aytar [4 ] introduce the concept of rough statistical cluster point in the following way: Let r r r be a non-negative real number. An element ζ \zeta ζ in a norm linear space is called a rough statistical cluster point w.r.t the roughness degree r r r (or, r r r -statistical cluster point) of a sequence x = { x n } n ∈ N x=\{x_n\}_{n\in\mathbb{N}} x = { x n } n ∈ N provided that
lim n → ∞ 1 n ∣ { k ≤ n : ∥ x k − ζ ∥ < r + ε } ∣ ≠ 0 , (2) \lim_{n\to\infty} \frac{1}{n}\left|\{k\le n : \lVert x_k-\zeta\rVert< r+\varepsilon\}\right|\ne0, \tag*{(2)} n → ∞ lim n 1 ∣ { k ≤ n : ∥ x k − ζ ∥ < r + ε } ∣ = 0 , (2)
for every ε > 0 \varepsilon>0 ε > 0 . The set of all r r r -statistical cluster points of the sequence x = { x n } n ∈ N x=\{x_n\}_{n\in\mathbb{N}} x = { x n } n ∈ N is symbolized by Γ x r \Gamma_x^r Γ x r . If r = 0 r=0 r = 0 in the above expression we could visualize r r r -statistical cluster point exactly to be merged with ordinary statistical cluster point [14 , 15 ] .
Theorem 1.1. [15 Proposition 2] For any sequence of real numbers x = { x n } n ∈ N x=\{x_n\}_{n\in\mathbb{N}} x = { x n } n ∈ N , the set Γ x \Gamma_x Γ x of statistical cluster points of the sequence x = { x n } n ∈ N x=\{x_n\}_{n\in\mathbb{N}} x = { x n } n ∈ N is a closed set.
Theorem 1.2. [11 Theorem 2.9] In a topological space the statistical cluster points set of a sequence x = { x n } n ∈ N x=\{x_n\}_{n\in\mathbb{N}} x = { x n } n ∈ N is closed.
Theorem 1.3. [4 Theorem 1.1] For any sequence x = { x n } n ∈ N x = \{x_n\}_{n\in\mathbb{N}} x = { x n } n ∈ N in a norm linear space X X X , the rough statistical cluster point set Γ x r \Gamma_x^r Γ x r is closed, for every r ≥ 0 r \ge0 r ≥ 0 .
Di Maio and Kočinac et al. [11 ] introduced the concept of statistical convergence in topology. Our aim is to introduce rough weighted ideal convergence and rough weighted ideal cluster points for a net in the context of topological vector spaces. In this context, we introduce the following definitions and results in sequence, which are primarily required.
Dündar et al. [12 , 13 ] and Pal et al. [27 ] independently introduced the concept of rough I \mathcal{I} I -convergence as a generalization of rough convergence [28 , 29 ] , rough statistical convergence [2 , 24 , 25 ] , and I \mathcal{I} I -convergence [9 , 20 , 21 , 33 ] . This notion is based on the structure of the ideal I \mathcal{I} I of subsets of the set N \mathbb{N} N . A family I ⊂ 2 N \mathcal{I} \subset2^{\mathbb{N}} I ⊂ 2 N is called an ideal on N \mathbb{N} N if (i) ∅ ∈ I \varnothing\in\mathcal{I} ∅ ∈ I (ii) A , B ∈ I A, B \in\mathcal{I} A , B ∈ I implies A ∪ B ∈ I A \cup B \in\mathcal{I} A ∪ B ∈ I (iii) A ∈ I , B ⊂ A A \in\mathcal{I}, B \subset A A ∈ I , B ⊂ A imply B ∈ I B \in\mathcal{I} B ∈ I . The ideal I \mathcal{I} I is called non-trivial if I ≠ { ∅ } \mathcal{I} \ne\{\varnothing\} I = { ∅ } and N ∉ I \mathbb{N} \notin\mathcal{I} N ∈ / I and a non-trivial ideal I \mathcal{I} I is called admissible if I \mathcal{I} I contains all the singleton set. Clearly I ⊂ 2 N \mathcal{I} \subset2^{\mathbb{N}} I ⊂ 2 N is a non-trivial ideal of N \mathbb{N} N iff F = F ( I ) = { K ⊂ N : N ∖ K ∈ I } \mathcal{F} = \mathcal{F}(\mathcal{I}) = \{K \subset\mathbb{N} : \mathbb{N} \setminus K \in\mathcal{I}\} F = F ( I ) = { K ⊂ N : N ∖ K ∈ I } is a filter on N \mathbb{N} N , called the filter associated with I \mathcal{I} I . For more details on the various types of rough convergence in recent work, see references [5 , 23 , 26 , 30 ] .
Definition 1.4. [12 , 13 , 27 ] Let r r r be a non-negative real number. A sequence x = { x n } n ∈ N x = \{x_n\}_{n\in\mathbb{N}} x = { x n } n ∈ N in a norm linear space is said to be rough I \mathcal{I} I -convergent to x ∗ x_* x ∗ w.r.t the roughness of degree r r r (where r ≥ 0 r \ge0 r ≥ 0 ), denoted by x n → I r x ∗ x_n \xrightarrow{\mathcal{I}}_r x_* x n I r x ∗ , provided for any ε > 0 \varepsilon> 0 ε > 0 , the set { n ∈ N : ∥ x n − x ∗ ∥ ≥ r + ε } ∈ I \{n \in\mathbb{N} : \|x_n - x_*\| \ge r + \varepsilon\} \in\mathcal{I} { n ∈ N : ∥ x n − x ∗ ∥ ≥ r + ε } ∈ I .
Definition 1.5. [7 , 21 ] Let ( X , d ) (X, d) ( X , d ) be a metric space. A point ζ ∈ X \zeta\in X ζ ∈ X is called an I \mathcal{I} I -cluster point of a sequence x = { x n } n ∈ N x = \{x_n\}_{n\in\mathbb{N}} x = { x n } n ∈ N in X X X if for any ε > 0 \varepsilon> 0 ε > 0 , the set { n ∈ N : d ( x n , ζ ) < ε } ∉ I \{n \in\mathbb{N} : d(x_n, \zeta) < \varepsilon\} \notin\mathcal{I} { n ∈ N : d ( x n , ζ ) < ε } ∈ / I .
The set of all I \mathcal{I} I -cluster point of x = { x n } n ∈ N x = \{x_n\}_{n\in\mathbb{N}} x = { x n } n ∈ N in a metric space ( X , d ) (X, d) ( X , d ) is denoted by Γ x ( I ) \Gamma_x(\mathcal{I}) Γ x ( I ) and the set Γ x ( I ) \Gamma_x(\mathcal{I}) Γ x ( I ) is closed in X X X , for any admissible ideal I \mathcal{I} I [6 , 7 , 21 ] .
Theorem 1.6. [7 ] Let I \mathcal{I} I be an admissible ideal on N \mathbb{N} N , Γ x ( I ) \Gamma_x(\mathcal{I}) Γ x ( I ) be the set of all I \mathcal{I} I -cluster points of a sequence x = { x n } n ∈ N x = \{x_n\}_{n\in\mathbb{N}} x = { x n } n ∈ N and K K K be a compact subset of a metric space X X X . Then for every sequence x = { x n } n ∈ N x = \{x_n\}_{n\in\mathbb{N}} x = { x n } n ∈ N in X X X with { n ∈ N : x n ∈ K } ∉ I \{n \in\mathbb{N} : x_n \in K\} \notin\mathcal{I} { n ∈ N : x n ∈ K } ∈ / I we have K ∩ Γ x ( I ) ≠ ∅ K \cap\Gamma_x(\mathcal{I}) \ne\varnothing K ∩ Γ x ( I ) = ∅ .
Lahiri et al. [22 ] initiate the idea of I \mathcal{I} I -convergence net in a topological space and establish several results using the concept. They define the I \mathcal{I} I -convergence net as follows: Let ( D , ≥ ) (D, \ge) ( D , ≥ ) be a directed set and α 0 ∈ D \alpha_0 \in D α 0 ∈ D , D α 0 = { α ∈ D : α ≥ α 0 } D_{\alpha_0} = \{\alpha\in D : \alpha\ge\alpha_0\} D α 0 = { α ∈ D : α ≥ α 0 } . An ideal I ⊂ 2 D \mathcal{I} \subset2^D I ⊂ 2 D is called non-trivial if I ≠ { ∅ } \mathcal{I} \ne\{\varnothing\} I = { ∅ } and D ∉ I D \notin\mathcal{I} D ∈ / I . A non-trivial ideal I \mathcal{I} I of D D D will be called D D D -admissible , if D α ∈ F ( I ) D_\alpha\in\mathcal{F}(\mathcal{I}) D α ∈ F ( I ) for all α ∈ D \alpha\in D α ∈ D . If I \mathcal{I} I is a non-trivial ideal of D D D , then F ( I ) = { M ⊂ D : M c ∈ I } \mathcal{F}(\mathcal{I}) = \{M \subset D : M^c \in\mathcal{I}\} F ( I ) = { M ⊂ D : M c ∈ I } is a filter associated with the ideal I \mathcal{I} I . For example the collection
F 0 = { A ⊂ D : there exists a D α 0 with A ⊃ D α 0 } \mathcal{F}_0 = \{A \subset D : \text{there exists a } D_{\alpha_0} \text{ with } A \supset D_{\alpha_0}\} F 0 = { A ⊂ D : there exists a D α 0 with A ⊃ D α 0 }
forms a filter in D D D . Therefore I 0 = { A ⊂ D : A c ∈ F 0 } \mathcal{I}_0 = \{A \subset D : A^c \in\mathcal{F}_0\} I 0 = { A ⊂ D : A c ∈ F 0 } is a non-trivial D D D -admissible ideal of D D D .
Definition 1.7. [22 ] Let I \mathcal{I} I be a non-trivial ideal of the directed set D D D . A net s = { s α } α ∈ D s = \{s_\alpha\}_{\alpha\in D} s = { s α } α ∈ D in an arbitrary topological space X X X is said to be I \mathcal{I} I -convergent to x 0 ∈ X x_0 \in X x 0 ∈ X , if for any open set U x 0 U_{x_0} U x 0 containing x 0 x_0 x 0 the set { α ∈ D : s α ∉ U x 0 } ∈ I \{\alpha\in D : s_\alpha\notin U_{x_0}\} \in\mathcal{I} { α ∈ D : s α ∈ / U x 0 } ∈ I . Symbolically we write I − lim s α = x 0 \mathcal{I} - \lim s_\alpha= x_0 I − lim s α = x 0 and we say that x 0 x_0 x 0 is the I \mathcal{I} I -limit of the net s = { s α } α ∈ D s = \{s_\alpha\}_{\alpha\in D} s = { s α } α ∈ D .
Definition 1.8. [22 ] Let I \mathcal{I} I be a non-trivial ideal of the directed set D D D and s = { s α } α ∈ D s = \{s_\alpha\}_{\alpha\in D} s = { s α } α ∈ D be a net in an arbitrary topological space X X X . A point ζ ∈ X \zeta\in X ζ ∈ X is called I \mathcal{I} I -cluster point of the net s = { s α } α ∈ D s = \{s_\alpha\}_{\alpha\in D} s = { s α } α ∈ D , if for any open set U ζ U_\zeta U ζ containing ζ \zeta ζ the set { α ∈ D : s α ∈ U ζ } ∉ I \{\alpha\in D : s_\alpha\in U_\zeta\} \notin\mathcal{I} { α ∈ D : s α ∈ U ζ } ∈ / I .
Theorem 1.9. [22 Theorem 9] Let I \mathcal{I} I be an ideal on a directed set D D D . In a topological space, the I \mathcal{I} I -cluster points set of a net is closed.
Definition 1.10. [22 ] A net ω = { ω e } e ∈ E \omega= \{\omega_e\}_{e\in E} ω = { ω e } e ∈ E , where E E E is a directed set, is said to be a subnet of the net s = { s α } α ∈ D s = \{s_\alpha\}_{\alpha\in D} s = { s α } α ∈ D if there is a mapping j : E → D j : E \to D j : E → D such that
(i): ω = s o i \omega= soi ω = so i
(ii): for any α ∈ D \alpha\in D α ∈ D there is an element e 0 ∈ E e_0 \in E e 0 ∈ E such that i ( e ) ≥ α i(e) \ge\alpha i ( e ) ≥ α for all e ∈ E e \in E e ∈ E with e ≥ e 0 e \ge e_0 e ≥ e 0 .
On the other hand, we briefly recall some of the basic definitions in the theory of Riesz space , and we refer the readers to [1 , 31 , 34 ] for more details.
Definition 1.12. [1 , 19 ] Consider L L L is a vector space over the field R \mathbb{R} R as well as a lattice with respect to the partial ordering ≤ \le ≤ on L L L . Then L L L is said to be a Riesz space (or a vector lattice ) if it satisfies the following properties:
(i) If a , b ∈ L a,b \in L a , b ∈ L and b ≤ a b \le a b ≤ a , then b + c ≤ a + c b+c \le a+c b + c ≤ a + c for each c ∈ L c \in L c ∈ L ,
(ii) If a , b ∈ L a,b \in L a , b ∈ L and b ≤ a b \le a b ≤ a , then λ b ≤ λ a \lambda b \le\lambda a λb ≤ λa for each λ ≥ 0 \lambda\ge0 λ ≥ 0 .
The absolute value of a ∈ L a \in L a ∈ L is defined by ∣ a ∣ = a ∨ ( − a ) |a| = a \vee(-a) ∣ a ∣ = a ∨ ( − a ) and a + = a ∨ θ a^{+} = a \vee\theta a + = a ∨ θ , a − = ( − a ) ∨ θ a^{-} = (-a) \vee\theta a − = ( − a ) ∨ θ , where θ \theta θ is the zero element of L L L . A subset A A A of a Riesz space L L L is said to be solid , if b ∈ A b \in A b ∈ A and ∣ a ∣ ≤ ∣ b ∣ |a| \le|b| ∣ a ∣ ≤ ∣ b ∣ imply a ∈ A a \in A a ∈ A . A topology τ \tau τ on L L L is a linear topology if the addition and the scalar multiplication are continuous. Every linear topology τ \tau τ on a vector space L L L has a base N \mathcal{N} N for the τ \tau τ -neighbourhoods of θ \theta θ (zero) satisfying the following properties:
(i): Each U θ ∈ N U_\theta\in\mathcal{N} U θ ∈ N is a balanced set ,
(ii): Each U θ ∈ N U_\theta\in\mathcal{N} U θ ∈ N is an absorbing set ,
(iii): For each U θ ∈ N U_\theta\in\mathcal{N} U θ ∈ N there exists some W θ ∈ N W_\theta\in\mathcal{N} W θ ∈ N with W θ + W θ ⊆ U θ W_\theta+ W_\theta\subseteq U_\theta W θ + W θ ⊆ U θ .
Definition 1.13. [1 , 19 ] A linear topology τ \tau τ on a Riesz space L L L is said to be locally solid if τ \tau τ has a base at θ \theta θ consisting of solid sets . A locally solid Riesz space ( L , τ ) (L,\tau) ( L , τ ) is a Riesz space L L L equipped with a locally solid topology τ \tau τ .
The symbol N s o l \mathcal{N}_{\mathrm{sol}} N sol stands for a base at θ \theta θ consisting of solid sets and satisfying the properties (i), (ii), and (iii) in a locally solid topology .
Definition 1.14. [16 , 19 ] A subset A A A of a Riesz space ( L , τ ) (L,\tau) ( L , τ ) is said to be τ \tau τ -bounded (or topologically bounded ) if for every τ \tau τ -neighbourhood U θ U_\theta U θ of θ \theta θ there exists some λ > 0 \lambda> 0 λ > 0 such that A ⊂ λ U θ A \subset\lambda U_\theta A ⊂ λ U θ . If A A A is not τ \tau τ -bounded , then it is called τ \tau τ -unbounded .
Definition 1.15. [10 , 19 ] Let ( L , τ ) (L,\tau) ( L , τ ) be a locally solid Riesz space . A net s = { s α } α ∈ D s = \{s_\alpha\}_{\alpha\in D} s = { s α } α ∈ D in L L L is said to be I τ I_\tau I τ -convergent (or, ideal-τ \tau τ -convergent ) to x 0 ∈ L x_0 \in L x 0 ∈ L if for any τ \tau τ -neighbourhood U θ U_\theta U θ of θ \theta θ , { α ∈ D : s α − x 0 ∉ U θ } ∈ I \{\alpha\in D : s_\alpha- x_0 \notin U_\theta\} \in I { α ∈ D : s α − x 0 ∈ / U θ } ∈ I . Briefly we write I τ − lim s α = x 0 I_\tau-\lim s_\alpha= x_0 I τ − lim s α = x 0 (or, s α → I τ x 0 s_\alpha\xrightarrow{I_\tau} x_0 s α I τ x 0 ).
Definition 1.16. [10 , 19 ] A net s = { s α } α ∈ D s = \{s_\alpha\}_{\alpha\in D} s = { s α } α ∈ D in a locally solid Riesz space ( L , τ ) (L,\tau) ( L , τ ) is said to be I τ I_\tau I τ -bounded (or, ideal-τ \tau τ -bounded ), if for any τ \tau τ -neighbourhood U θ U_\theta U θ of θ \theta θ there exists some λ > 0 \lambda> 0 λ > 0 such that { α ∈ D : λ s α ∉ U θ } ∈ I \{\alpha\in D : \lambda s_\alpha\notin U_\theta\} \in I { α ∈ D : λ s α ∈ / U θ } ∈ I .
In this discourse, we now proceed to the definitions of rough weighted I τ I_\tau I τ -convergence [18 ] and weighted I τ I_\tau I τ -cluster point [18 ] of a sequence in a locally solid Riesz space ( L , τ ) (L,\tau) ( L , τ ) . Meanwhile, we define the weighted sequence and weighted convergence sequence of real numbers.
Definition 1.17. [16 , 18 , 19 ] A sequence of real numbers t = { t n } n ∈ N t = \{t_n\}_{n\in\mathbb{N}} t = { t n } n ∈ N is said to be weighted sequence if there exists a positive real number δ \delta δ such that
t n > δ for all n ∈ N . (3) t_n > \delta\quad\text{for all } n \in\mathbb{N}. \tag*{(3)} t n > δ for all n ∈ N . (3)
Definition 1.18. [16 , 17 ] A sequence of real numbers { x n } n ∈ N \{x_n\}_{n\in\mathbb{N}} { x n } n ∈ N is said to converge in a weighted manner to ℓ \ell ℓ with respect to the weighted sequence t = { t n } n ∈ N t = \{t_n\}_{n\in\mathbb{N}} t = { t n } n ∈ N if, for every ε > 0 \varepsilon> 0 ε > 0 , there exists a positive integer k k k such that
t n ∣ x n − ℓ ∣ < ε holds for all n ≥ k . (4) t_n|x_n-\ell| < \varepsilon\text{ holds for all } n \ge k. \tag*{(4)} t n ∣ x n − ℓ ∣ < ε holds for all n ≥ k . (4)
This relationship is expressed as x n → W t ℓ x_n \xrightarrow{W^t} \ell x n W t ℓ .
Remark 1.19. If the weighted sequence t = { t n } n ∈ N t=\{t_n\}_{n\in\mathbb{N}} t = { t n } n ∈ N does not satisfy the required condition of (3) , i.e.,
t n > δ for all n ∈ N , where δ > 0 is a constant, (5) t_n>\delta\text{ for all } n\in\mathbb{N},\text{ where }\delta>0\text{ is a constant,} \tag*{(5)} t n > δ for all n ∈ N , where δ > 0 is a constant, (5)
a significant problem may occur. For example, let lim t n = 0 \lim t_n=0 lim t n = 0 . From (4) , we can conclude that any bounded sequence x = { x n } n ∈ N x=\{x_n\}_{n\in\mathbb{N}} x = { x n } n ∈ N is weighted convergent to any real number. However, this conclusion lacks meaningful interpretation.
If a sequence is convergent in the usual sense, it may not be a weighted convergent sequence. The following example gives the required answer.
Example 1.20. Let { x n } n ∈ N \{x_n\}_{n\in\mathbb{N}} { x n } n ∈ N be a sequence of real numbers such that x n = 1 n x_n=\frac{1}{n} x n = n 1 for all n ∈ N n\in\mathbb{N} n ∈ N , and a weighted sequence be t n = n 2 t_n=n^2 t n = n 2 for all n ∈ N n\in\mathbb{N} n ∈ N . Then, t n ∣ x n − 0 ∣ = n → ∞ t_n|x_n-0|=n\to\infty t n ∣ x n − 0∣ = n → ∞ as n → ∞ n\to\infty n → ∞ . Therefore, x n → W t 0 x_n \xrightarrow{W^t} 0 x n W t 0 . In conclusion, we say that the usual and weighted convergence are two entirely different notions.
Definition 1.21. [18 ] Let I \mathcal{I} I be an admissible ideal of N \mathbb{N} N and t = { t n } n ∈ N t=\{t_n\}_{n\in\mathbb{N}} t = { t n } n ∈ N be a weighted sequence of real numbers. A sequence x = { x n } n ∈ N x=\{x_n\}_{n\in\mathbb{N}} x = { x n } n ∈ N in a locally solid Riesz space ( L , τ ) (L,\tau) ( L , τ ) is said to be rough weighted I τ \mathcal{I}_{\tau} I τ -convergent to x ∗ ∈ L x_*\in L x ∗ ∈ L w.r.t the roughness of degree V θ V_{\theta} V θ (where V θ V_{\theta} V θ is a τ \tau τ -neighbourhood of θ \theta θ ), denoted by x n → V θ W t x ∗ x_n \xrightarrow[\ V_{\theta}\ ]{W^t} x_* x n W t V θ x ∗ , if for every τ \tau τ -neighbourhood U θ U_{\theta} U θ of θ \theta θ , the following expression holds,
{ n ∈ N : t n ( x n − x ∗ ) ∉ V θ + U θ } ∈ I \{n\in\mathbb{N}:t_n(x_n-x_*)\notin V_{\theta}+U_{\theta}\}\in\mathcal{I} { n ∈ N : t n ( x n − x ∗ ) ∈ / V θ + U θ } ∈ I or, { n ∈ N : t n ( x n − x ∗ ) ∈ V θ + U θ } ∈ F ( I ) \{n\in\mathbb{N}:t_n(x_n-x_*)\in V_{\theta}+U_{\theta}\}\in\mathcal{F}(\mathcal{I}) { n ∈ N : t n ( x n − x ∗ ) ∈ V θ + U θ } ∈ F ( I ) .
We shall write W I τ - L I M V θ x = { x ∗ ∈ L : x n → V θ W t x ∗ } \mathrm{WI}_{\tau}\text{-}\mathrm{LIM}^{V_{\theta}}x=\{x_*\in L:x_n\xrightarrow[\ V_{\theta}\ ]{W^t}x_*\} WI τ - LIM V θ x = { x ∗ ∈ L : x n W t V θ x ∗ } to denote the set of all rough weighted I τ \mathcal{I}_{\tau} I τ -limit points of the sequence x = { x n } n ∈ N x=\{x_n\}_{n\in\mathbb{N}} x = { x n } n ∈ N with degree of roughness V θ V_{\theta} V θ .
Definition 1.22. [18 ] Let I \mathcal{I} I be an admissible ideal of N \mathbb{N} N and t = { t n } n ∈ N t=\{t_n\}_{n\in\mathbb{N}} t = { t n } n ∈ N be a weighted sequence of real numbers. An element ζ ∈ L \zeta\in L ζ ∈ L is called weighted I τ \mathcal{I}_{\tau} I τ -cluster point of a sequence x = { x n } n ∈ N x=\{x_n\}_{n\in\mathbb{N}} x = { x n } n ∈ N in L L L if for every τ \tau τ -neighbourhood U θ U_{\theta} U θ of θ \theta θ , the set
{ n ∈ N : t n ( x n − ζ ) ∈ U θ } ∉ I \{n\in\mathbb{N}:t_n(x_n-\zeta)\in U_{\theta}\}\notin\mathcal{I} { n ∈ N : t n ( x n − ζ ) ∈ U θ } ∈ / I .
The set of all weighted I τ \mathcal{I}_{\tau} I τ -cluster points of the sequence x = { x n } n ∈ N x=\{x_n\}_{n\in\mathbb{N}} x = { x n } n ∈ N is denoted by W I τ Γ x \mathrm{WI}_{\tau}\Gamma_x WI τ Γ x .
Remark 1.23. In the Definition 1.17 [16 , 18 , 19 ] , a positive real number δ \delta δ is defined in the context of weighted sequence t = { t n } n ∈ N t=\{t_n\}_{n\in\mathbb{N}} t = { t n } n ∈ N , such that t n > δ t_n>\delta t n > δ for every n ∈ N n\in\mathbb{N} n ∈ N . We do not follow the usual convention and propose a new hypothesis in the Definition 1.24, as follows:
there exists a δ > 0 \delta>0 δ > 0 such that ∣ t α ∣ > δ |t_{\alpha}|>\delta ∣ t α ∣ > δ for all α ∈ D \alpha\in D α ∈ D , where D D D is a directed set.
Definition 1.24. Let ( D , ≥ ) (D,\geq) ( D , ≥ ) be a directed set. A net t = { t α } α ∈ D t=\{t_{\alpha}\}_{\alpha\in D} t = { t α } α ∈ D in R \mathbb{R} R is said to be a weighted net, if there exists a positive real number δ \delta δ such that ∣ t α ∣ > δ |t_{\alpha}|>\delta ∣ t α ∣ > δ for all α ∈ D \alpha\in D α ∈ D .
Continuing from our previous discussion, we now present two important definitions.
Definition 1.25. Let I \mathcal{I} I be a non-trivial ideal of the directed set D D D . A weighted net t = { t α } α ∈ D t=\{t_{\alpha}\}_{\alpha\in D} t = { t α } α ∈ D is said to be I \mathcal{I} I -bounded, if there exists a positive real number M M M such that
{ α ∈ D : ∣ t α ∣ > M } ∈ I \{\alpha\in D:|t_{\alpha}|>M\}\in\mathcal{I} { α ∈ D : ∣ t α ∣ > M } ∈ I or, { α ∈ D : ∣ t α ∣ ≤ M } ∈ F ( I ) \{\alpha\in D:|t_{\alpha}|\leq M\}\in\mathcal{F}(\mathcal{I}) { α ∈ D : ∣ t α ∣ ≤ M } ∈ F ( I ) .
Definition 1.26. Let ( L , τ ) (L,\tau) ( L , τ ) be a locally solid Riesz space and I \mathcal{I} I be a non-trivial ideal of the directed set D D D . A net s = { s α } α ∈ D s=\{s_{\alpha}\}_{\alpha\in D} s = { s α } α ∈ D in L L L is said to be weighted I τ \mathcal{I}_{\tau} I τ -bounded corresponding to the weighted net t = { t α } α ∈ D t=\{t_{\alpha}\}_{\alpha\in D} t = { t α } α ∈ D if for every τ \tau τ -neighbourhood U θ U_{\theta} U θ of θ \theta θ there exists a λ > 0 \lambda>0 λ > 0 such that
{ α ∈ D : λ t α s α ∉ U θ } ∈ I \{\alpha\in D:\lambda t_{\alpha}s_{\alpha}\notin U_{\theta}\}\in\mathcal{I} { α ∈ D : λ t α s α ∈ / U θ } ∈ I or, { α ∈ D : λ t α s α ∈ U θ } ∈ F ( I ) \{\alpha\in D:\lambda t_{\alpha}s_{\alpha}\in U_{\theta}\}\in\mathcal{F}(\mathcal{I}) { α ∈ D : λ t α s α ∈ U θ } ∈ F ( I ) .
Main Results
Let us begin this section by asking how to define ‘rough weighted I τ \mathcal{I}_{\tau} I τ -convergence’ and ‘rough weighted I τ \mathcal{I}_{\tau} I τ -cluster point’ of a net in locally solid Riesz spaces. From the Equations 1.1, 1.2, we redefine the concept of ‘rough statistical convergence’ and ‘rough statistical cluster point’ of a sequence in a normed linear space X X X using neighbourhoods of θ \theta θ , as follows:
Definition 2.1. Let X X X be a normed linear space, B ( θ , M ) = { x ∈ X : ∥ x − θ ∥ < M } B(\theta,M)=\{x\in X:\lVert x-\theta\rVert<M\} B ( θ , M ) = { x ∈ X : ∥ x − θ ∥ < M } where M > 0 M>0 M > 0 and
B ( θ , M 1 ) + B ( θ , M 2 ) = { x 1 + x 2 ∈ X : x 1 ∈ B ( θ , M 1 ) and x 2 ∈ B ( θ , M 2 ) } . B(\theta,M_1)+B(\theta,M_2)=\{x_1+x_2\in X:x_1\in B(\theta,M_1)\text{ and }x_2\in B(\theta,M_2)\}. B ( θ , M 1 ) + B ( θ , M 2 ) = { x 1 + x 2 ∈ X : x 1 ∈ B ( θ , M 1 ) and x 2 ∈ B ( θ , M 2 )} . (i): A sequence x = { x n } n ∈ N x=\{x_n\}_{n\in\mathbb{N}} x = { x n } n ∈ N in X X X is said to be rough statistically convergent to x ∗ x_* x ∗ w.r.t the roughness degree r > 0 r>0 r > 0 , if for any ε > 0 \varepsilon>0 ε > 0 , such that
lim n → ∞ 1 n ∣ { k ≤ n : x k − x ∗ ∉ B ( θ , r ) + B ( θ , ε ) } ∣ = 0 , (6) \lim_{n\to\infty}\frac{1}{n}\left|\{k\le n:x_k-x_*\notin B(\theta,r)+B(\theta,\varepsilon)\}\right|=0, \tag*{(6)} n → ∞ lim n 1 ∣ { k ≤ n : x k − x ∗ ∈ / B ( θ , r ) + B ( θ , ε )} ∣ = 0 , (6) (ii): An element ζ \zeta ζ in a norm space X X X is called a rough statistical cluster point w.r.t the roughness degree r > 0 r>0 r > 0 of sequence x = { x n } n ∈ N x=\{x_n\}_{n\in\mathbb{N}} x = { x n } n ∈ N if for any ε > 0 \varepsilon>0 ε > 0 , such that
lim n → ∞ 1 n ∣ { k ≤ n : x k − ζ ∈ B ( θ , r ) + B ( θ , ε ) } ∣ ≠ 0. (7) \lim_{n\to\infty}\frac{1}{n}\left|\{k\le n:x_k-\zeta\in B(\theta,r)+B(\theta,\varepsilon)\}\right|\ne0. \tag*{(7)} n → ∞ lim n 1 ∣ { k ≤ n : x k − ζ ∈ B ( θ , r ) + B ( θ , ε )} ∣ = 0. (7)
According to the Equations 2.1 and 2.2, we are now able to introduce the definitions of ‘rough weighted I τ \mathcal{I}_{\tau} I τ -convergence’ and ‘rough weighted I τ \mathcal{I}_{\tau} I τ -cluster point’ of a net in locally solid Riesz spaces as follows:
Definition 2.2. Let I \mathcal{I} I be a non-trivial ideal of a directed set D D D and t = { t α } α ∈ D t=\{t_\alpha\}_{\alpha\in D} t = { t α } α ∈ D be a weighted net. A net s = { s α } α ∈ D s=\{s_\alpha\}_{\alpha\in D} s = { s α } α ∈ D in a locally solid Riesz space ( L , τ ) (L,\tau) ( L , τ ) is said to be rough weighted I τ \mathcal{I}_{\tau} I τ -convergent to s 0 ∈ L s_0\in L s 0 ∈ L w.r.t the roughness of degree V θ V_\theta V θ (where V θ V_\theta V θ is a τ \tau τ -neighbourhood of θ \theta θ ), denoted by s α → V θ W I τ s 0 s_\alpha\xrightarrow[ V_\theta]{\mathrm{W}\mathcal{I}_{\tau}}s_0 s α W I τ V θ s 0 , if for every τ \tau τ -neighbourhood U θ U_\theta U θ of θ \theta θ , the following expression holds,
{ α ∈ D : t α ( s α − s 0 ) ∉ V θ + U θ } ∈ I or { α ∈ D : t α ( s α − s 0 ) ∈ V θ + U θ } ∈ F ( I ) . \{\alpha\in D:t_\alpha(s_\alpha-s_0)\notin V_\theta+U_\theta\}\in\mathcal{I}\text{ or }\{\alpha\in D:t_\alpha(s_\alpha-s_0)\in V_\theta+U_\theta\}\in\mathcal{F}(\mathcal{I}). { α ∈ D : t α ( s α − s 0 ) ∈ / V θ + U θ } ∈ I or { α ∈ D : t α ( s α − s 0 ) ∈ V θ + U θ } ∈ F ( I ) . We shall write W I τ - L I M I V θ s = { s 0 ∈ L : s α → V θ W I τ s 0 } \mathrm{W}\mathcal{I}_{\tau}\text{-}\mathrm{LIM}^{V_\theta}_{\mathcal{I}}s=\{s_0\in L:s_\alpha\xrightarrow[ V_\theta]{\mathrm{W}\mathcal{I}_{\tau}}s_0\} W I τ - LIM I V θ s = { s 0 ∈ L : s α W I τ V θ s 0 } to denote the set of all rough weighted I τ \mathcal{I}_{\tau} I τ -limit points of the net s = { s α } α ∈ D s=\{s_\alpha\}_{\alpha\in D} s = { s α } α ∈ D with degree of roughness V θ V_\theta V θ .
Definition 2.3. Let I \mathcal{I} I be a non-trivial ideal of a directed set D D D , L L L be a locally solid Riesz space and t = { t α } α ∈ D t=\{t_\alpha\}_{\alpha\in D} t = { t α } α ∈ D be a weighted net. A point ζ ∈ L \zeta\in L ζ ∈ L is called a rough weighted I τ \mathcal{I}_{\tau} I τ -cluster point of the net s = { s α } α ∈ D s=\{s_\alpha\}_{\alpha\in D} s = { s α } α ∈ D w.r.t the roughness of degree V θ V_\theta V θ (where V θ V_\theta V θ is a τ \tau τ -neighbourhood of θ \theta θ ), if for every τ \tau τ -neighbourhood U θ U_\theta U θ of θ \theta θ ,
{ α ∈ D : t α ( s α − ζ ) ∈ V θ + U θ } ∉ I . \{\alpha\in D:t_\alpha(s_\alpha-\zeta)\in V_\theta+U_\theta\}\notin\mathcal{I}. { α ∈ D : t α ( s α − ζ ) ∈ V θ + U θ } ∈ / I . The set of all rough weighted I τ \mathcal{I}_{\tau} I τ -cluster points of a net s = { s α } α ∈ D s=\{s_\alpha\}_{\alpha\in D} s = { s α } α ∈ D is denoted by $\mathcal{R}^{V_\theta}_{t}-\mathcal{I}(\Gamma_s^{\tau}).
Example 2.5. (i) Let us consider the locally solid Riesz space ( R 2 , ∥ ⋅ ∥ ) (\mathbb{R}^{2},\lVert\cdot\rVert) ( R 2 , ∥ ⋅ ∥) with the Euclidean norm ∥ ⋅ ∥ \lVert\cdot\rVert ∥ ⋅ ∥ and coordinate-wise ordering. The family N s o l \mathcal{N}_{sol} N so l of all U ( r ) U(r) U ( r ) is defined as U θ ( r ) = { x ∈ R 2 : ∥ x ∥ < r } = { ( x , y ) ∈ R 2 : x 2 + y 2 < r 2 } U_{\theta}(r)=\{x\in\mathbb{R}^{2}:\lVert x\rVert<r\}=\{(x,y)\in\mathbb{R}^{2}:x^{2}+y^{2}<r^{2}\} U θ ( r ) = { x ∈ R 2 : ∥ x ∥ < r } = {( x , y ) ∈ R 2 : x 2 + y 2 < r 2 } where r > 0 r>0 r > 0 , is a base at θ = ( 0 , 0 ) \theta=(0,0) θ = ( 0 , 0 ) . It is known that N s o l \mathcal{N}_{sol} N so l is itself a directed set D D D with the binary relation ≥ \geq ≥ defined by U θ ≥ V θ U_{\theta}\geq V_{\theta} U θ ≥ V θ if U θ ⊇ V θ U_{\theta}\supseteq V_{\theta} U θ ⊇ V θ for any U θ , V θ ∈ D U_{\theta},V_{\theta}\in D U θ , V θ ∈ D . Let W θ ∈ D = N s o l W_{\theta}\in D=\mathcal{N}_{sol} W θ ∈ D = N so l , D W θ = { Z θ ∈ D : Z θ ≥ W θ } D_{W_{\theta}}=\{Z_{\theta}\in D:Z_{\theta}\geq W_{\theta}\} D W θ = { Z θ ∈ D : Z θ ≥ W θ } , then the collection
F 0 = { A ⊂ D : there exists a set D W θ with A ⊃ D W θ } \mathcal{F}_{0}=\{A\subset D:\text{ there exists a set }D_{W_{\theta}}\text{ with }A\supset D_{W_{\theta}}\} F 0 = { A ⊂ D : there exists a set D W θ with A ⊃ D W θ } forms a filter in D D D . Let I 0 = { A ⊂ D : A c ∈ F 0 } \mathcal{I}_{0}=\{A\subset D:A^{c}\in\mathcal{F}_{0}\} I 0 = { A ⊂ D : A c ∈ F 0 } . Then I 0 \mathcal{I}_{0} I 0 is a non-trivial ideal of D D D . Also G = { U θ ( r ) : r ∈ R + ∖ Q + } G=\{U_{\theta}(r):r\in\mathbb{R}^{+}\setminus\mathbb{Q}^{+}\} G = { U θ ( r ) : r ∈ R + ∖ Q + } neither belongs to I 0 \mathcal{I}_{0} I 0 nor to F 0 \mathcal{F}_{0} F 0 .
Next we choose a net s = { s U θ } U θ ∈ D s=\{s_{U_{\theta}}\}_{U_{\theta}\in D} s = { s U θ } U θ ∈ D , a weighted net t = { t U θ } U θ ∈ D t=\{t_{U_{\theta}}\}_{U_{\theta}\in D} t = { t U θ } U θ ∈ D and a τ \tau τ -neighbourhood of θ \theta θ , say V θ V_{\theta} V θ , in the following manner:
s U θ ( r ) ∈ { ( 1 , 0 ) + { U θ ( 1 ) ∖ U θ ( 1 ) } if U θ ( r ) ∈ G , ( 4 , 0 ) + { U θ ( 4 ) ∖ U θ ( 4 ) } if U θ ( r ) ∈ D ∖ G , and t U θ ( r ) = { e r 1 + e r if r ∈ G , r + 1 if r ∈ D ∖ G , s_{U_{\theta}(r)}\in
\begin{cases}
(1,0)+\{U_{\theta}(1)\setminus U_{\theta}(1)\} & \text{if }U_{\theta}(r)\in G,\\
(4,0)+\{U_{\theta}(4)\setminus U_{\theta}(4)\} & \text{if }U_{\theta}(r)\in D\setminus G,
\end{cases}
\qquad\text{and}\qquad
t_{U_{\theta}(r)}=
\begin{cases}
\dfrac{e^{r}}{1+e^{r}} & \text{if }r\in G,\\
r+1 & \text{if }r\in D\setminus G,
\end{cases} s U θ ( r ) ∈ { ( 1 , 0 ) + { U θ ( 1 ) ∖ U θ ( 1 )} ( 4 , 0 ) + { U θ ( 4 ) ∖ U θ ( 4 )} if U θ ( r ) ∈ G , if U θ ( r ) ∈ D ∖ G , and t U θ ( r ) = ⎩ ⎨ ⎧ 1 + e r e r r + 1 if r ∈ G , if r ∈ D ∖ G , i.e.,
s U θ ( r ) ∈ { ( 1 , 0 ) + { ( x , y ) ∈ R 2 : x 2 + y 2 = 1 } if U θ ( r ) ∈ G , ( 4 , 0 ) + { ( x , y ) ∈ R 2 : x 2 + y 2 = 4 2 } if U θ ( r ) ∈ D ∖ G , s_{U_{\theta}(r)}\in
\begin{cases}
(1,0)+\{(x,y)\in\mathbb{R}^{2}:x^{2}+y^{2}=1\} & \text{if }U_{\theta}(r)\in G,\\
(4,0)+\{(x,y)\in\mathbb{R}^{2}:x^{2}+y^{2}=4^{2}\} & \text{if }U_{\theta}(r)\in D\setminus G,
\end{cases} s U θ ( r ) ∈ { ( 1 , 0 ) + {( x , y ) ∈ R 2 : x 2 + y 2 = 1 } ( 4 , 0 ) + {( x , y ) ∈ R 2 : x 2 + y 2 = 4 2 } if U θ ( r ) ∈ G , if U θ ( r ) ∈ D ∖ G , and V θ = U θ ( 1 ) = { ( x , y ) ∈ R 2 : x 2 + y 2 ≤ 1 } V_{\theta}=U_{\theta}(1)=\{(x,y)\in\mathbb{R}^{2}:x^{2}+y^{2}\leq1\} V θ = U θ ( 1 ) = {( x , y ) ∈ R 2 : x 2 + y 2 ≤ 1 } . Then
{ U θ ( r ) ∈ D : t U θ ( r ) ( s U θ ( r ) − ( 1 , 0 ) ) ∈ V θ + W θ } = G \{U_{\theta}(r)\in D:t_{U_{\theta}(r)}(s_{U_{\theta}(r)}-(1,0))\in V_{\theta}+W_{\theta}\}=G { U θ ( r ) ∈ D : t U θ ( r ) ( s U θ ( r ) − ( 1 , 0 )) ∈ V θ + W θ } = G neither belongs to I 0 \mathcal{I}_{0} I 0 nor to F 0 \mathcal{F}_{0} F 0 , for any τ \tau τ -neighbourhood W θ W_{\theta} W θ of θ \theta θ where W θ ⊂ V θ W_{\theta}\subset V_{\theta} W θ ⊂ V θ . This implies ( 1 , 0 ) ∈ R t V θ − I 0 ( Γ s τ ) (1,0)\in R_{t}^{V^{\theta}}-\mathcal{I}_{0}(\Gamma_{s}^{\tau}) ( 1 , 0 ) ∈ R t V θ − I 0 ( Γ s τ ) , whereas the net s = { s U θ } U θ ∈ D s=\{s_{U_{\theta}}\}_{U_{\theta}\in D} s = { s U θ } U θ ∈ D cannot rough weighted I τ \mathcal{I}_{\tau} I τ -converge to ( 1 , 0 ) (1,0) ( 1 , 0 ) w.r.t the roughness of degree V θ V_{\theta} V θ .
( 1 , 0 ) ∈ ( R t V θ − I ( Γ s τ ) ) ∖ ( W I τ − LIM t V θ s ) ≠ ∅ . (1,0)\in\left(R_{t}^{V^{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau})\right)\setminus\left(\operatorname{W}\mathcal{I}_{\tau}-\operatorname{LIM}_{t}^{V^{\theta}}s\right)\ne\varnothing. ( 1 , 0 ) ∈ ( R t V θ − I ( Γ s τ ) ) ∖ ( W I τ − LIM t V θ s ) = ∅ . (ii) Let { A n : n ∈ N } \{A_{n}:n\in\mathbb{N}\} { A n : n ∈ N } be a partition of the directed set D = R + D=\mathbb{R}^{+} D = R + into infinite sets as in Example 2.9 (i). Also, V θ = ( − 1 2 , 1 2 ) V_{\theta}=(-\frac{1}{2},\frac{1}{2}) V θ = ( − 2 1 , 2 1 ) and U θ ⊂ V θ U_{\theta}\subset V_{\theta} U θ ⊂ V θ . For each n ∈ N n\in\mathbb{N} n ∈ N and α ∈ A n \alpha\in A_{n} α ∈ A n , define s α = n s_{\alpha}=n s α = n and t α = n t_{\alpha}=n t α = n . Let I ⊂ P ( R + ) \mathcal{I}\subset\mathcal{P}(\mathbb{R}^{+}) I ⊂ P ( R + ) be the ideal of sets I I I satisfying that I ∩ A n I\cap A_{n} I ∩ A n is finite for all n ∈ N n\in\mathbb{N} n ∈ N . Then the set { α ∈ R + : t α ( s α − 1 ) = 0 ∈ V θ + U θ } = A 1 \{\alpha\in\mathbb{R}^{+}:t_{\alpha}(s_{\alpha}-1)=0\in V_{\theta}+U_{\theta}\}=A_{1} { α ∈ R + : t α ( s α − 1 ) = 0 ∈ V θ + U θ } = A 1 neither belongs to I \mathcal{I} I nor to F ( I ) \mathcal{F}(\mathcal{I}) F ( I ) . This implies that
1 ∈ ( R t V θ − I ( Γ s τ ) ) ∖ ( W I τ − LIM t V θ s ) ≠ ∅ . 1\in\left(R_{t}^{V^{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau})\right)\setminus\left(\operatorname{W}\mathcal{I}_{\tau}-\operatorname{LIM}_{t}^{V^{\theta}}s\right)\ne\varnothing. 1 ∈ ( R t V θ − I ( Γ s τ ) ) ∖ ( W I τ − LIM t V θ s ) = ∅ .
We now prove the expected relationship between the sets W I τ − LIM t V θ s \operatorname{W}\mathcal{I}_{\tau}-\operatorname{LIM}_{t}^{V^{\theta}}s W I τ − LIM t V θ s and R t V θ − I ( Γ τ ) R_{t}^{V^{\theta}}-\mathcal{I}(\Gamma^{\tau}) R t V θ − I ( Γ τ ) .
Theorem 2.6. Let I D \mathcal{I}^{D} I D be an D D D -admissible ideal of a directed set D D D , t = { t α } α ∈ D t=\{t_{\alpha}\}_{\alpha\in D} t = { t α } α ∈ D be a weighted net , s = { s α } α ∈ D s=\{s_{\alpha}\}_{\alpha\in D} s = { s α } α ∈ D be a net in a locally solid Riesz space L L L and V θ V_{\theta} V θ be a τ \tau τ -neighbourhood of θ \theta θ . If ζ ∈ R t V θ − I D ( Γ s τ ) \zeta\in R_{t}^{V^{\theta}}-\mathcal{I}^{D}(\Gamma_{s}^{\tau}) ζ ∈ R t V θ − I D ( Γ s τ ) , then there exists a subnet ω = { ω ε } ε ∈ E \omega=\{\omega_{\varepsilon}\}_{\varepsilon\in E} ω = { ω ε } ε ∈ E of the net { t α ( s α − ζ ) } α ∈ D \{t_{\alpha}(s_{\alpha}-\zeta)\}_{\alpha\in D} { t α ( s α − ζ ) } α ∈ D such that θ ∈ W I τ E − LIM 1 W θ ω \theta\in\operatorname{W}\mathcal{I}_{\tau}^{E}-\operatorname{LIM}_{1}^{W_{\theta}}\omega θ ∈ W I τ E − LIM 1 W θ ω , where I E \mathcal{I}^{E} I E is a E E E -admissible ideal of E E E and W θ W_{\theta} W θ is τ \tau τ -neighbourhood of θ \theta θ such that V θ ⊆ W θ V_{\theta}\subseteq W_{\theta} V θ ⊆ W θ .
Proof. Since ζ ∈ R t V θ − I D ( Γ τ ) \zeta\in R_t^{V_\theta}-\mathcal{I}^{D}(\Gamma^\tau) ζ ∈ R t V θ − I D ( Γ τ ) , then for every τ \tau τ -neighbourhood U θ U_\theta U θ of θ \theta θ , the set
A = { α ∈ D : t α ( s α − ζ ) ∈ V θ + U θ } ∉ I D A=\{\alpha\in D:t_\alpha(s_\alpha-\zeta)\in V_\theta+U_\theta\}\notin\mathcal{I}^{D} A = { α ∈ D : t α ( s α − ζ ) ∈ V θ + U θ } ∈ / I D
⟹ D ∖ A = { α ∈ D : t α ( s α − ζ ) ∉ V θ + U θ } ∉ F ( I D ) . \Longrightarrow D\setminus A=\{\alpha\in D:t_\alpha(s_\alpha-\zeta)\notin V_\theta+U_\theta\}\notin\mathcal{F}(\mathcal{I}^{D}). ⟹ D ∖ A = { α ∈ D : t α ( s α − ζ ) ∈ / V θ + U θ } ∈ / F ( I D ) .
Let α 0 ∈ D \alpha_0\in D α 0 ∈ D . Then D α 0 = { α ∈ D : α ≥ α 0 } ∈ F ( I D ) D_{\alpha_0}=\{\alpha\in D:\alpha\geq\alpha_0\}\in\mathcal{F}(\mathcal{I}^{D}) D α 0 = { α ∈ D : α ≥ α 0 } ∈ F ( I D ) and D α 0 ⊈ D ∖ A D_{\alpha_0}\not\subseteq D\setminus A D α 0 ⊆ D ∖ A (otherwise D ∖ A ∈ F ( I D ) D\setminus A\in\mathcal{F}(\mathcal{I}^{D}) D ∖ A ∈ F ( I D ) , a contradiction). There is an element α ∗ ∈ D α 0 \alpha_*\in D_{\alpha_0} α ∗ ∈ D α 0 such that α ∗ ∉ D ∖ A \alpha_*\notin D\setminus A α ∗ ∈ / D ∖ A . This implies α ∗ ∈ A ⊂ D \alpha_*\in A\subset D α ∗ ∈ A ⊂ D , t α ∗ ( s α ∗ − ζ ) ∈ V θ + U θ t_{\alpha_*}(s_{\alpha_*}-\zeta)\in V_\theta+U_\theta t α ∗ ( s α ∗ − ζ ) ∈ V θ + U θ and α ∗ ≥ α 0 \alpha_*\geq\alpha_0 α ∗ ≥ α 0 . So α ∗ \alpha_* α ∗ depending on α 0 \alpha_0 α 0 and U θ U_\theta U θ . We construct a directed set E E E , a subnet ω = { ω e } e ∈ E \omega=\{\omega_e\}_{e\in E} ω = { ω e } e ∈ E of the net { t α ( s α − ζ ) } α ∈ D \{t_\alpha(s_\alpha-\zeta)\}_{\alpha\in D} { t α ( s α − ζ ) } α ∈ D as follows:
E = { ( α , U θ ) ∈ D × N θ : t α ( s α − ζ ) ∈ V θ + U θ } , E=\{(\alpha,U_\theta)\in D\times\mathcal{N}_\theta:t_\alpha(s_\alpha-\zeta)\in V_\theta+U_\theta\}, E = {( α , U θ ) ∈ D × N θ : t α ( s α − ζ ) ∈ V θ + U θ } ,
where N θ \mathcal{N}_\theta N θ denote the family of all τ \tau τ -neighbourhood of θ \theta θ . For ( α , U θ ) , ( β , U θ ∗ ) ∈ E (\alpha,U_\theta),(\beta,U_\theta^*)\in E ( α , U θ ) , ( β , U θ ∗ ) ∈ E define a relation ( β , U θ ∗ ) ≥ ( α , U θ ) (\beta,U_\theta^*)\geq(\alpha,U_\theta) ( β , U θ ∗ ) ≥ ( α , U θ ) mean β ≥ α \beta\geq\alpha β ≥ α and U θ ∗ ⊆ U θ U_\theta^*\subseteq U_\theta U θ ∗ ⊆ U θ . Then E E E is a directed set. Now we choose two mappings i : E → D i:E\to D i : E → D and ω : E → L \omega:E\to L ω : E → L as follows:
i ( ( α 0 , U θ ) ) = α ∗ and ω ( ( α 0 , U θ ) ) = t α ∗ ( s α ∗ − ζ ) for all ( α 0 , U θ ) ∈ E . i((\alpha_0,U_\theta))=\alpha_*\quad\text{and}\quad\omega((\alpha_0,U_\theta))=t_{\alpha_*}(s_{\alpha_*}-\zeta)\quad\text{for all }(\alpha_0,U_\theta)\in E. i (( α 0 , U θ )) = α ∗ and ω (( α 0 , U θ )) = t α ∗ ( s α ∗ − ζ ) for all ( α 0 , U θ ) ∈ E .
Let U θ ∗ U_\theta^* U θ ∗ be a τ \tau τ -neighbourhood of θ \theta θ and β ∈ D \beta\in D β ∈ D . Then there exists an element β ∗ ∈ D \beta_*\in D β ∗ ∈ D depends on U θ ∗ U_\theta^* U θ ∗ and β \beta β with β ∗ ≥ β \beta_*\geq\beta β ∗ ≥ β such that
ω ( ( β , U θ ∗ ) ) = t β ∗ ( s β ∗ − ζ ) ∈ V θ + U θ ∗ ⊆ V θ + U θ ⊆ W θ + U θ , \omega((\beta,U_\theta^*))=t_{\beta_*}(s_{\beta_*}-\zeta)\in V_\theta+U_\theta^*\subseteq V_\theta+U_\theta\subseteq W_\theta+U_\theta, ω (( β , U θ ∗ )) = t β ∗ ( s β ∗ − ζ ) ∈ V θ + U θ ∗ ⊆ V θ + U θ ⊆ W θ + U θ ,
for all ( β , U θ ∗ ) ≥ ( α 0 , U θ ) (\beta,U_\theta^*)\geq(\alpha_0,U_\theta) ( β , U θ ∗ ) ≥ ( α 0 , U θ ) , β ≥ α 0 \beta\geq\alpha_0 β ≥ α 0 and U θ ∗ ⊆ U θ U_\theta^*\subseteq U_\theta U θ ∗ ⊆ U θ . Therefore
E ( α 0 , U θ ) ⊆ { ( β , U θ ∗ ) ∈ E : ω ( ( β , U θ ∗ ) ) ∈ W θ + U θ } ∈ F ( I E ) . E_{(\alpha_0,U_\theta)}\subseteq\{(\beta,U_\theta^*)\in E:\omega((\beta,U_\theta^*))\in W_\theta+U_\theta\}\in\mathcal{F}(\mathcal{I}^{E}). E ( α 0 , U θ ) ⊆ {( β , U θ ∗ ) ∈ E : ω (( β , U θ ∗ )) ∈ W θ + U θ } ∈ F ( I E ) .
Therefore this is the result.
Theorem 2.7. Let D D D be a directed set, t = { t α } α ∈ D t=\{t_\alpha\}_{\alpha\in D} t = { t α } α ∈ D be a weighted net, s = { s α } α ∈ D s=\{s_\alpha\}_{\alpha\in D} s = { s α } α ∈ D be a net in a locally solid Riesz space L L L , ζ ∈ L \zeta\in L ζ ∈ L , W θ W_\theta W θ be a τ \tau τ -neighbourhood of θ \theta θ and ω = { ω e } e ∈ E \omega=\{\omega_e\}_{e\in E} ω = { ω e } e ∈ E be a subnet of the net { t α ( s α − ζ ) } α ∈ D \{t_\alpha(s_\alpha-\zeta)\}_{\alpha\in D} { t α ( s α − ζ ) } α ∈ D . If θ ∈ W I τ E − LIM 1 W θ ω \theta\in WI_\tau^{E}-\operatorname{LIM}_1^{W_\theta}\omega θ ∈ W I τ E − LIM 1 W θ ω , where I E \mathcal{I}^{E} I E is a E E E -admissible ideal of E E E then ζ ∈ R t V θ − I D ( Γ τ ) \zeta\in R_t^{V_\theta}-\mathcal{I}^{D}(\Gamma^\tau) ζ ∈ R t V θ − I D ( Γ τ ) , for some D D D -admissible ideal I D \mathcal{I}^{D} I D and W θ ⊆ V θ W_\theta\subseteq V_\theta W θ ⊆ V θ .
Proof. Since ω = { ω e } e ∈ E \omega=\{\omega_e\}_{e\in E} ω = { ω e } e ∈ E is a subnet of the net { t α ( s α − ζ ) } α ∈ D \{t_\alpha(s_\alpha-\zeta)\}_{\alpha\in D} { t α ( s α − ζ ) } α ∈ D there exists a mapping i : E → D i:E\to D i : E → D such that ω ( e ) = S ∘ i ( e ) \omega(e)=S\circ i(e) ω ( e ) = S ∘ i ( e ) for all e ∈ E e\in E e ∈ E , where S α = t α ( s α − ζ ) S_\alpha=t_\alpha(s_\alpha-\zeta) S α = t α ( s α − ζ ) for all α ∈ D \alpha\in D α ∈ D . As in Remark 1.11, the collection I D = { A ⊂ D : i − 1 ( A ) ∈ I E } \mathcal{I}^{D}=\{A\subset D:i^{-1}(A)\in\mathcal{I}^{E}\} I D = { A ⊂ D : i − 1 ( A ) ∈ I E } is a D D D -admissible ideal of D D D . We assume by the way of contradiction ζ ∉ R t V θ − I D ( Γ τ ) \zeta\notin R_t^{V_\theta}-\mathcal{I}^{D}(\Gamma^\tau) ζ ∈ / R t V θ − I D ( Γ τ ) , then the set
A = { α ∈ D : t α ( s α − ζ ) ∈ V θ + U θ } ∈ I D A=\{\alpha\in D:t_\alpha(s_\alpha-\zeta)\in V_\theta+U_\theta\}\in\mathcal{I}^{D} A = { α ∈ D : t α ( s α − ζ ) ∈ V θ + U θ } ∈ I D and i − 1 ( A ) ∈ I E i^{-1}(A)\in\mathcal{I}^{E} i − 1 ( A ) ∈ I E .
Since θ ∈ W I τ E − LIM 1 W θ ω \theta\in WI_\tau^{E}-\operatorname{LIM}_1^{W_\theta}\omega θ ∈ W I τ E − LIM 1 W θ ω , then the set
C = { e ∈ E : ω e ∈ W θ + U θ } ∈ F ( I E ) . C=\{e\in E:\omega_e\in W_\theta+U_\theta\}\in\mathcal{F}(\mathcal{I}^{E}). C = { e ∈ E : ω e ∈ W θ + U θ } ∈ F ( I E ) . Let e ∈ C e\in C e ∈ C . Then
ω e ∈ W θ + U θ ⟹ S ∘ i ( e ) ∈ W θ + U θ ⟹ S ( i ( e ) ) ∈ W θ + U θ ⊆ V θ + U θ , \omega_e\in W_\theta+U_\theta\Longrightarrow S\circ i(e)\in W_\theta+U_\theta\Longrightarrow S(i(e))\in W_\theta+U_\theta\subseteq V_\theta+U_\theta, ω e ∈ W θ + U θ ⟹ S ∘ i ( e ) ∈ W θ + U θ ⟹ S ( i ( e )) ∈ W θ + U θ ⊆ V θ + U θ , ⟹ i ( e ) ∈ A ⟹ e ∈ i − 1 ( A ) . \Longrightarrow i(e)\in A\Longrightarrow e\in i^{-1}(A). ⟹ i ( e ) ∈ A ⟹ e ∈ i − 1 ( A ) . Therefore C ⊆ i − 1 ( A ) C\subseteq i^{-1}(A) C ⊆ i − 1 ( A ) . Since C ∈ F ( I E ) C\in\mathcal{F}(\mathcal{I}^{E}) C ∈ F ( I E ) so i − 1 ( A ) ∈ F ( I E ) i^{-1}(A)\in\mathcal{F}(\mathcal{I}^{E}) i − 1 ( A ) ∈ F ( I E ) , which is a contradiction since i − 1 ( A ) ∈ I E i^{-1}(A)\in\mathcal{I}^{E} i − 1 ( A ) ∈ I E .
As an immediate consequence of the above two theorems we get following remark.
Recall an important result of Lahiri et al. [22 Theorem 9] ‘In a topological space the I \mathcal{I} I -cluster points set of a net is closed’. While reinventing this theorem in terms of ‘rough weighted I τ \mathcal{I}_\tau I τ -cluster points set of a net in a locally solid Riesz space’, we assert that rough weighted I τ \mathcal{I}_\tau I τ -cluster points set of a net in a locally solid Riesz space is not necessarily closed. Our affirmation is emphatically demonstrated in the following example.
Example 2.9. (i) Consider the locally solid Riesz space ( R 2 , ∥ ⋅ ∥ ) (\mathbb{R}^{2},\lVert\cdot\rVert) ( R 2 , ∥ ⋅ ∥) in conjunction with the Euclidean norm ∥ ⋅ ∥ \lVert\cdot\rVert ∥ ⋅ ∥ and coordinate-wise ordering. The directed set D D D of all U θ ( r ) U_{\theta}(r) U θ ( r ) is defined as
U θ ( r ) = { x ∈ R 2 : ∥ x ∥ < r } = { ( x , y ) ∈ R 2 : x 2 + y 2 < r 2 } U_{\theta}(r)=\{x\in\mathbb{R}^{2}:\lVert x\rVert<r\}=\{(x,y)\in\mathbb{R}^{2}:x^{2}+y^{2}<r^{2}\} U θ ( r ) = { x ∈ R 2 : ∥ x ∥ < r } = {( x , y ) ∈ R 2 : x 2 + y 2 < r 2 } where r > 0 r>0 r > 0 with the binary relation ≥ \geq ≥ defined by U θ ≥ V θ U_{\theta}\geq V_{\theta} U θ ≥ V θ if U θ ⊆ V θ U_{\theta}\subseteq V_{\theta} U θ ⊆ V θ for any U θ , V θ ∈ D U_{\theta},V_{\theta}\in D U θ , V θ ∈ D . Consider a new decomposition of R + \mathbb{R}^{+} R + ,
B 0 = ( 0 , 1 ) ∪ ⋃ n ∈ O [ n , n + 1 ] B_{0}=(0,1)\cup\bigcup_{n\in O}[n,n+1] B 0 = ( 0 , 1 ) ∪ n ∈ O ⋃ [ n , n + 1 ] where O = { 1 , 3 , 5 , … } O=\{1,3,5,\ldots\} O = { 1 , 3 , 5 , … } ,
B ℓ = ⋃ n ∈ O ( 2 ℓ n , 2 ℓ n + 1 ) B_{\ell}=\bigcup_{n\in O}(2^{\ell}n,2^{\ell}n+1) B ℓ = n ∈ O ⋃ ( 2 ℓ n , 2 ℓ n + 1 ) where ℓ ∈ N = { 1 , 2 , 3 , … } \ell\in\mathbb{N}=\{1,2,3,\ldots\} ℓ ∈ N = { 1 , 2 , 3 , … } ,
⟹ R + = ⋃ ℓ ∈ N 0 B ℓ \Longrightarrow\mathbb{R}^{+}=\bigcup_{\ell\in\mathbb{N}_{0}}B_{\ell} ⟹ R + = ℓ ∈ N 0 ⋃ B ℓ where B i ∩ B j = ∅ B_{i}\cap B_{j}=\varnothing B i ∩ B j = ∅ for all i , j ∈ N 0 = { 0 , 1 , 2 , 3 , … } i,j\in\mathbb{N}_{0}=\{0,1,2,3,\ldots\} i , j ∈ N 0 = { 0 , 1 , 2 , 3 , … } and i ≠ j i\ne j i = j .
Let I = { A ⊂ D : A is countable } \mathcal{I}=\{A\subset D:A\text{ is countable}\} I = { A ⊂ D : A is countable } be a proper non-trivial ideal of D D D and V θ = U θ ( 1 ) ‾ = { ( x , y ) ∈ R 2 : x 2 + y 2 ≤ 1 } V_{\theta}=\overline{U_{\theta}(1)}=\{(x,y)\in\mathbb{R}^{2}:x^{2}+y^{2}\leq1\} V θ = U θ ( 1 ) = {( x , y ) ∈ R 2 : x 2 + y 2 ≤ 1 } . Now we select a net s = { s α } α ∈ D s=\{s_{\alpha}\}_{\alpha\in D} s = { s α } α ∈ D and a weighted net t = { t α } α ∈ D t=\{t_{\alpha}\}_{\alpha\in D} t = { t α } α ∈ D in the following manner:
s U θ ( r ) = { ( 2 + r , 1 ) if r ∈ B 0 , ( 1 2 ℓ + 1 r + 1 r 2 , 1 ) if r ∈ B ℓ , ℓ ∈ N , and t U θ ( r ) = { 2 if r ∈ B 0 , r + 1 if r ∈ B ℓ , ℓ ∈ N . s_{U_{\theta}(r)}=
\begin{cases}
(2+r,1)&\text{if }r\in B_{0},\\
\left(\frac{1}{2^{\ell}}+\frac{1}{r}+\frac{1}{r^{2}},1\right)&\text{if }r\in B_{\ell},\ \ell\in\mathbb{N},
\end{cases}
\qquad\text{and}\qquad
t_{U_{\theta}(r)}=
\begin{cases}
2&\text{if }r\in B_{0},\\
r+1&\text{if }r\in B_{\ell},\ \ell\in\mathbb{N}.
\end{cases} s U θ ( r ) = { ( 2 + r , 1 ) ( 2 ℓ 1 + r 1 + r 2 1 , 1 ) if r ∈ B 0 , if r ∈ B ℓ , ℓ ∈ N , and t U θ ( r ) = { 2 r + 1 if r ∈ B 0 , if r ∈ B ℓ , ℓ ∈ N . Let U θ U_{\theta} U θ be a τ \tau τ -neighbourhood of θ \theta θ , there exists some U θ ( ε ) ∈ N s o l U_{\theta}(\varepsilon)\in\mathcal{N}_{\mathrm{sol}} U θ ( ε ) ∈ N sol , 0 < ε < 1 2 0<\varepsilon<\frac{1}{2} 0 < ε < 2 1 such that U θ ( ε ) = { ( x , y ) ∈ R 2 : x 2 + y 2 < ε 2 } ⊂ U θ U_{\theta}(\varepsilon)=\{(x,y)\in\mathbb{R}^{2}:x^{2}+y^{2}<\varepsilon^{2}\}\subset U_{\theta} U θ ( ε ) = {( x , y ) ∈ R 2 : x 2 + y 2 < ε 2 } ⊂ U θ . Moreover for each m ∈ N m\in\mathbb{N} m ∈ N , we derive that
t U θ ( r ) ( s U θ ( r ) − ( 1 2 m , 1 ) ) = { ( 4 + 2 r − 1 2 m − 1 , 0 ) if r ∈ B 0 , ( ( r + 1 ) ( 1 2 ℓ − 1 2 m ) + 1 + 2 r + 1 r 2 , 0 ) if r ∈ B ℓ , ℓ ∈ N . t_{U_{\theta}(r)}\left(s_{U_{\theta}(r)}-\left(\frac{1}{2^{m}},1\right)\right)=
\begin{cases}
\left(4+2r-\frac{1}{2^{m-1}},0\right)&\text{if }r\in B_{0},\\
\left((r+1)\left(\frac{1}{2^{\ell}}-\frac{1}{2^{m}}\right)+1+\frac{2}{r}+\frac{1}{r^{2}},0\right)&\text{if }r\in B_{\ell},\ \ell\in\mathbb{N}.
\end{cases} t U θ ( r ) ( s U θ ( r ) − ( 2 m 1 , 1 ) ) = { ( 4 + 2 r − 2 m − 1 1 , 0 ) ( ( r + 1 ) ( 2 ℓ 1 − 2 m 1 ) + 1 + r 2 + r 2 1 , 0 ) if r ∈ B 0 , if r ∈ B ℓ , ℓ ∈ N . Let r ∈ B ℓ r\in B_{\ell} r ∈ B ℓ , ℓ ∈ N \ell\in\mathbb{N} ℓ ∈ N and m = ℓ m=\ell m = ℓ then
t U θ ( r ) ( s U θ ( r ) − ( 1 2 m , 1 ) ) = ( 1 + 2 r + 1 r 2 , 0 ) ∈ V θ + U θ ( ε ) , t_{U_{\theta}(r)}\left(s_{U_{\theta}(r)}-\left(\frac{1}{2^{m}},1\right)\right)=\left(1+\frac{2}{r}+\frac{1}{r^{2}},0\right)\in V_{\theta}+U_{\theta}(\varepsilon), t U θ ( r ) ( s U θ ( r ) − ( 2 m 1 , 1 ) ) = ( 1 + r 2 + r 2 1 , 0 ) ∈ V θ + U θ ( ε ) , ⟹ { U θ ( r ) ∈ D : t U θ ( r ) ( s U θ ( r ) − ( 1 2 m , 1 ) ) ∈ V θ + U θ ( ε ) } ∉ I . \Longrightarrow\{U_{\theta}(r)\in D:t_{U_{\theta}(r)}(s_{U_{\theta}(r)}-(\frac{1}{2^{m}},1))\in V_{\theta}+U_{\theta}(\varepsilon)\}\notin\mathcal{I}. ⟹ { U θ ( r ) ∈ D : t U θ ( r ) ( s U θ ( r ) − ( 2 m 1 , 1 )) ∈ V θ + U θ ( ε )} ∈ / I . This conveys ( 1 2 m , 1 ) ∈ R t V θ − I ( Γ s τ ) (\frac{1}{2^{m}},1)\in R_{t}^{V_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau}) ( 2 m 1 , 1 ) ∈ R t V θ − I ( Γ s τ ) for all m ∈ N m\in\mathbb{N} m ∈ N . For r ∈ B ℓ r\in B_{\ell} r ∈ B ℓ where ℓ ∈ N \ell\in\mathbb{N} ℓ ∈ N there exists an odd positive integer n 0 ∈ O n_{0}\in O n 0 ∈ O such that r ∈ ( 2 ℓ n 0 , 2 ℓ n 0 + 1 ) r\in(2^{\ell}n_{0},2^{\ell}n_{0}+1) r ∈ ( 2 ℓ n 0 , 2 ℓ n 0 + 1 ) then r > 2 ℓ n 0 r>2^{\ell}n_{0} r > 2 ℓ n 0 implies r + 1 2 ℓ ≥ n 0 ≥ 1 \frac{r+1}{2^{\ell}}\geq n_{0}\geq1 2 ℓ r + 1 ≥ n 0 ≥ 1 . Hence we get
{ U θ ( r ) ∈ D : t U θ ( r ) ( s U θ ( r ) − ( 0 , 1 ) ) ∈ V θ + U θ ( ε ) } \{U_{\theta}(r)\in D:t_{U_{\theta}(r)}(s_{U_{\theta}(r)}-(0,1))\in V_{\theta}+U_{\theta}(\varepsilon)\} { U θ ( r ) ∈ D : t U θ ( r ) ( s U θ ( r ) − ( 0 , 1 )) ∈ V θ + U θ ( ε )} = { { U θ ( r ) ∈ D , r ∈ B 0 : ( 4 + 2 r , 0 ) ∈ V θ + U θ ( ε ) } = { Finite set } ∈ I , { U θ ( r ) ∈ D , r ∈ B ℓ for ℓ ∈ N : ( 1 + r + 1 r + 2 r + 1 r 2 , 0 ) ∈ V θ + U θ ( ε ) } = { Finite set } ∈ I . =\left\{
\begin{array}{l}
\{U_{\theta}(r)\in D,r\in B_{0}:(4+2r,0)\in V_{\theta}+U_{\theta}(\varepsilon)\}=\{\text{Finite set}\}\in\mathcal{I},\\
\{U_{\theta}(r)\in D,r\in B_{\ell}\text{ for }\ell\in\mathbb{N}:(1+\frac{r+1}{r}+\frac{2}{r}+\frac{1}{r^{2}},0)\in V_{\theta}+U_{\theta}(\varepsilon)\}=\{\text{Finite set}\}\in\mathcal{I}.
\end{array}
\right. = { { U θ ( r ) ∈ D , r ∈ B 0 : ( 4 + 2 r , 0 ) ∈ V θ + U θ ( ε )} = { Finite set } ∈ I , { U θ ( r ) ∈ D , r ∈ B ℓ for ℓ ∈ N : ( 1 + r r + 1 + r 2 + r 2 1 , 0 ) ∈ V θ + U θ ( ε )} = { Finite set } ∈ I . Above illustration emphasizes ( 0 , 1 ) ∉ R t V θ − I ( Γ s τ ) (0,1)\notin R_{t}^{V_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau}) ( 0 , 1 ) ∈ / R t V θ − I ( Γ s τ ) and so our affirmation R t V θ − I ( Γ s τ ) R_{t}^{V_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau}) R t V θ − I ( Γ s τ ) is not a closed set is established.
(ii) Let { x n } n ∈ N ⊂ ( 0 , 1 ) \{x_{n}\}_{n\in\mathbb{N}}\subset(0,1) { x n } n ∈ N ⊂ ( 0 , 1 ) be a sequence converging to 0 0 0 and { A n : n ∈ N } \{A_{n}:n\in\mathbb{N}\} { A n : n ∈ N } be a partition of D = R + D=\mathbb{R}^{+} D = R + into infinite sets as above. Also, V θ = ( − 1 , 1 ) V_{\theta}=(-1,1) V θ = ( − 1 , 1 ) and U θ ⊂ V θ U_{\theta}\subset V_{\theta} U θ ⊂ V θ . For each n ∈ N n\in\mathbb{N} n ∈ N and α ∈ A n \alpha\in A_{n} α ∈ A n , define s α = x n s_{\alpha}=x_{n} s α = x n and t α = 3 x n t_{\alpha}=\frac{3}{x_{n}} t α = x n 3 . Let I ⊂ P ( R + ) \mathcal{I}\subset\mathcal{P}(\mathbb{R}^{+}) I ⊂ P ( R + ) be the ideal of sets I I I satisfying that I ∩ A n I\cap A_{n} I ∩ A n is finite for all n ∈ N n\in\mathbb{N} n ∈ N . Then, for each n n n , the set { α ∈ R + : t α ( s α − x n ) = 0 ∈ V θ + U θ } \{\alpha\in\mathbb{R}^{+}:t_{\alpha}(s_{\alpha}-x_{n})=0\in V_{\theta}+U_{\theta}\} { α ∈ R + : t α ( s α − x n ) = 0 ∈ V θ + U θ } contains the set A n A_{n} A n and so is not in I \mathcal{I} I . This implies that all the point of the sequence { x n } n ∈ N \{x_{n}\}_{n\in\mathbb{N}} { x n } n ∈ N is a rough weighted I τ \mathcal{I}_{\tau} I τ -cluster point of the net s = { s α } α ∈ D s=\{s_{\alpha}\}_{\alpha\in D} s = { s α } α ∈ D i.e., x n ∈ R t V θ − I ( Γ s τ ) x_{n}\in R_{t}^{V_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau}) x n ∈ R t V θ − I ( Γ s τ ) for all n ∈ N n\in\mathbb{N} n ∈ N . On the other hand, the set { α ∈ R + : t α ( s α − 0 ) = 3 ∈ V θ + U θ } \{\alpha\in\mathbb{R}^{+}:t_{\alpha}(s_{\alpha}-0)=3\in V_{\theta}+U_{\theta}\} { α ∈ R + : t α ( s α − 0 ) = 3 ∈ V θ + U θ } is empty, and so 0 0 0 is not rough weighted I τ \mathcal{I}_{\tau} I τ -cluster point of the net s = { s α } α ∈ D s=\{s_{\alpha}\}_{\alpha\in D} s = { s α } α ∈ D i.e., 0 ∉ R t V θ − I ( Γ s τ ) 0\notin R_{t}^{V_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau}) 0 ∈ / R t V θ − I ( Γ s τ ) . Therefore, set R t V θ − I ( Γ s τ ) R_{t}^{V_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau}) R t V θ − I ( Γ s τ ) is not closed.
An important theorem is now presented, derived from the above example. We show the following results holds without using the locally solid Riesz spaces are first countable, Hausdorff or, V θ V_{\theta} V θ is τ \tau τ -bounded.
Proposition 2.10. Let I \mathcal{I} I be a non-trivial ideal of a directed set D D D and ( L , τ ) (L,\tau) ( L , τ ) be a locally solid Riesz space. If the weighted net t = { t α } α ∈ D t=\{t_{\alpha}\}_{\alpha\in D} t = { t α } α ∈ D is I \mathcal{I} I -bounded, then the set R t V θ − I ( Γ s τ ) \mathcal{R}_{t}^{V_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau}) R t V θ − I ( Γ s τ ) is closed.
Proof. From the I \mathcal{I} I -boundedness of the weighted net , there exists a positive real number M M M such that the set A = { α ∈ D : ∣ t α ∣ < M } ∈ F ( I ) A=\{\alpha\in D:|t_{\alpha}|<M\}\in\mathcal{F}(\mathcal{I}) A = { α ∈ D : ∣ t α ∣ < M } ∈ F ( I ) . If U θ U_{\theta} U θ be any neighbourhood of θ \theta θ , then there exists a W θ ∈ N s o l W_{\theta}\in\mathcal{N}_{sol} W θ ∈ N so l such that W θ + W θ ⊆ U θ W_{\theta}+W_{\theta}\subseteq U_{\theta} W θ + W θ ⊆ U θ . Let ω ∈ R t V θ − I ( Γ s τ ) \omega\in\mathcal{R}_{t}^{V_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau}) ω ∈ R t V θ − I ( Γ s τ ) . Then there exists a
ζ ∈ { R t V θ − I ( Γ s τ ) } ∩ ( ω + 1 M W θ ) . \zeta\in\{\mathcal{R}_{t}^{V_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau})\}\cap\left(\omega+\frac{1}{M}W_{\theta}\right). ζ ∈ { R t V θ − I ( Γ s τ )} ∩ ( ω + M 1 W θ ) .
This implies ( ζ − ω ) M ∈ W θ (\zeta-\omega)M\in W_{\theta} ( ζ − ω ) M ∈ W θ and B = { α ∈ D : t α ( s α − ζ ) ∈ V θ + W θ } ∉ I B=\{\alpha\in D:t_{\alpha}(s_{\alpha}-\zeta)\in V_{\theta}+W_{\theta}\}\notin\mathcal{I} B = { α ∈ D : t α ( s α − ζ ) ∈ V θ + W θ } ∈ / I . Therefore A ∩ B ∉ I A\cap B\notin\mathcal{I} A ∩ B ∈ / I (otherwise, B ⊆ ( D ∖ A ) ∪ ( A ∩ B ) ∈ I B\subseteq(D\setminus A)\cup(A\cap B)\in\mathcal{I} B ⊆ ( D ∖ A ) ∪ ( A ∩ B ) ∈ I , a contradiction). Let α ∈ A ∩ B \alpha\in A\cap B α ∈ A ∩ B then
t α ( s α − ω ) = t α ( s α − ζ ) + t α ( ζ − ω ) ∈ V θ + W θ + W θ ⊆ V θ + U θ , t_{\alpha}(s_{\alpha}-\omega)=t_{\alpha}(s_{\alpha}-\zeta)+t_{\alpha}(\zeta-\omega)\in V_{\theta}+W_{\theta}+W_{\theta}\subseteq V_{\theta}+U_{\theta}, t α ( s α − ω ) = t α ( s α − ζ ) + t α ( ζ − ω ) ∈ V θ + W θ + W θ ⊆ V θ + U θ ,
since ∣ t α ( ζ − ω ) ∣ ≤ ∣ M ( ζ − ω ) ∣ |t_{\alpha}(\zeta-\omega)|\leq|M(\zeta-\omega)| ∣ t α ( ζ − ω ) ∣ ≤ ∣ M ( ζ − ω ) ∣ , ( ζ − ω ) M ∈ W θ (\zeta-\omega)M\in W_{\theta} ( ζ − ω ) M ∈ W θ and W θ ∈ N s o l W_{\theta}\in\mathcal{N}_{sol} W θ ∈ N so l then t α ( ζ − ω ) ∈ W θ t_{\alpha}(\zeta-\omega)\in W_{\theta} t α ( ζ − ω ) ∈ W θ . Therefore
A ∩ B ⊆ { α ∈ D : t α ( s α − ω ) ∈ V θ + U θ } ∉ I . A\cap B\subseteq\{\alpha\in D:t_{\alpha}(s_{\alpha}-\omega)\in V_{\theta}+U_{\theta}\}\notin\mathcal{I}. A ∩ B ⊆ { α ∈ D : t α ( s α − ω ) ∈ V θ + U θ } ∈ / I .
Hence the set R t V θ − I ( Γ s τ ) \mathcal{R}_{t}^{V_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau}) R t V θ − I ( Γ s τ ) is closed.
The following results that shows non-emptiness of the set R t V θ − I ( Γ s τ ) \mathcal{R}_{t}^{V_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau}) R t V θ − I ( Γ s τ ) .
Proposition 2.11. If the nets s = { s α } α ∈ D s=\{s_{\alpha}\}_{\alpha\in D} s = { s α } α ∈ D is weighted I τ \mathcal{I}_{\tau} I τ -bounded corresponding to the weighted net t = { t α } α ∈ D t=\{t_{\alpha}\}_{\alpha\in D} t = { t α } α ∈ D , then for every τ \tau τ -neighbourhood U θ U_{\theta} U θ of θ \theta θ , there exists a positive real number ϑ \vartheta ϑ such that R t ϑ U θ − I ( Γ s τ ) ≠ ∅ \mathcal{R}_{t}^{\vartheta U_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau})\ne\varnothing R t ϑ U θ − I ( Γ s τ ) = ∅ .
Proof. Since s = { s α } α ∈ D s=\{s_{\alpha}\}_{\alpha\in D} s = { s α } α ∈ D is weighted I τ \mathcal{I}_{\tau} I τ -bounded then for every τ \tau τ -neighborhood U θ U_{\theta} U θ of θ \theta θ there exists some 1 ϑ > 0 \frac{1}{\vartheta}>0 ϑ 1 > 0 such that
{ α ∈ D : t α ( s α − θ ) ∈ ϑ U θ } = { α ∈ D : t α s α ∈ ϑ U θ } = { α ∈ D : 1 ϑ t α s α ∈ U θ } ∈ F ( I ) . \{\alpha\in D:t_{\alpha}(s_{\alpha}-\theta)\in\vartheta U_{\theta}\}=\{\alpha\in D:t_{\alpha}s_{\alpha}\in\vartheta U_{\theta}\}=\{\alpha\in D:\frac{1}{\vartheta}t_{\alpha}s_{\alpha}\in U_{\theta}\}\in\mathcal{F}(\mathcal{I}). { α ∈ D : t α ( s α − θ ) ∈ ϑ U θ } = { α ∈ D : t α s α ∈ ϑ U θ } = { α ∈ D : ϑ 1 t α s α ∈ U θ } ∈ F ( I ) .
Therefore θ ∈ R t ϑ U θ − I ( Γ s τ ) \theta\in\mathcal{R}_{t}^{\vartheta U_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau}) θ ∈ R t ϑ U θ − I ( Γ s τ ) . As a result, this is what we have concluded □
Next, we choose an example showing that the converse is not valid.
Example 2.12. Consider the locally solid Riesz space , directed set D D D , G G G and ideal I \mathcal{I} I as defined in Example 2.5 (i). Also we select the net s = { s α } α ∈ D s=\{s_{\alpha}\}_{\alpha\in D} s = { s α } α ∈ D and the weighted net t = { t α } α ∈ D t=\{t_{\alpha}\}_{\alpha\in D} t = { t α } α ∈ D in respectivly:
s U θ ( r ) = { ( 2 , 3 ) if U θ ( r ) ∈ D ∖ G , ( r , r ) if U θ ( r ) ∈ G , and t U θ ( r ) = − ( r + 1 ) for all U θ ( r ) ∈ D . s_{U_{\theta}(r)}=
\begin{cases}
(2,3) & \text{if }U_{\theta}(r)\in D\setminus G,\\
(r,r) & \text{if }U_{\theta}(r)\in G,
\end{cases}
\qquad\text{and}\qquad
t_{U_{\theta}(r)}=-(r+1)\text{ for all }U_{\theta}(r)\in D. s U θ ( r ) = { ( 2 , 3 ) ( r , r ) if U θ ( r ) ∈ D ∖ G , if U θ ( r ) ∈ G , and t U θ ( r ) = − ( r + 1 ) for all U θ ( r ) ∈ D . If U θ U_{\theta} U θ is any arbitrary τ \tau τ -neighbourhood of θ \theta θ and ϑ \vartheta ϑ is any positive real number, then ( 2 , 3 ) ∈ R t ϑ U θ − I ( Γ s τ ) (2,3)\in\mathcal{R}_{t}^{\vartheta U_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau}) ( 2 , 3 ) ∈ R t ϑ U θ − I ( Γ s τ ) and
{ U θ ( r ) ∈ D : ϑ t U θ ( r ) s U θ ( r ) ∉ U θ ( 1 ) } ∉ I . \{U_{\theta}(r)\in D:\vartheta t_{U_{\theta}(r)}s_{U_{\theta}(r)}\notin U_{\theta}(1)\}\notin\mathcal{I}. { U θ ( r ) ∈ D : ϑ t U θ ( r ) s U θ ( r ) ∈ / U θ ( 1 )} ∈ / I . Hence the nets s = { s α } α ∈ D s=\{s_{\alpha}\}_{\alpha\in D} s = { s α } α ∈ D is not weighted I τ \mathcal{I}_{\tau} I τ -bounded whereas R t ϑ U θ − I ( Γ s τ ) ≠ ∅ \mathcal{R}_{t}^{\vartheta U_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau})\ne\varnothing R t ϑ U θ − I ( Γ s τ ) = ∅ .
Now, we figure out some other results related to the non-emptiness of the set R t V θ − I ( Γ s τ ) \mathcal{R}_{t}^{V_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau}) R t V θ − I ( Γ s τ ) .
Proposition 2.13. Let I \mathcal{I} I be a non-trivial ideal of a directed set D D D , t = { t α } α ∈ D t=\{t_{\alpha}\}_{\alpha\in D} t = { t α } α ∈ D be a weighted net and s = { s α } α ∈ D s=\{s_{\alpha}\}_{\alpha\in D} s = { s α } α ∈ D be a net in a locally solid Riesz space L L L . Suppose H ℓ = { t α ( s α − ℓ ) ∈ L : α ∈ H } H_{\ell}=\{t_{\alpha}(s_{\alpha}-\ell)\in L:\alpha\in H\} H ℓ = { t α ( s α − ℓ ) ∈ L : α ∈ H } where ℓ ∈ L \ell\in L ℓ ∈ L and H ∈ F ( I ) H\in\mathcal{F}(\mathcal{I}) H ∈ F ( I ) .
ζ ∈ R t V θ − I ( Γ s τ ) \zeta\in\mathcal{R}_{t}^{V_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau}) ζ ∈ R t V θ − I ( Γ s τ ) if θ ∈ H ζ ‾ \theta\in\overline{H_{\zeta}} θ ∈ H ζ for all H ∈ F ( I ) H\in\mathcal{F}(\mathcal{I}) H ∈ F ( I ) ,
H ζ ∩ V θ ‾ ≠ ∅ H_{\zeta}\cap\overline{V_{\theta}}\ne\varnothing H ζ ∩ V θ = ∅ for all H ∈ F ( I ) H\in\mathcal{F}(\mathcal{I}) H ∈ F ( I ) , if ζ ∈ R t V θ − I ( Γ s τ ) \zeta\in\mathcal{R}_{t}^{V_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau}) ζ ∈ R t V θ − I ( Γ s τ ) and L L L is Hausdorff.
Proof. (i) If possible, let ζ \zeta ζ be not a rough weighted I τ I_{\tau} I τ -cluster point of s = { s α } α ∈ D s=\{s_{\alpha}\}_{\alpha\in D} s = { s α } α ∈ D . Then there exists a τ \tau τ -neighbourhood U θ \mathbf{U}_{\theta} U θ of θ \theta θ such that
H 0 = { α ∈ D : t α ( s α − ζ ) ∈ L ∖ ( V θ + U θ ) } ∈ F ( I ) , H^{0}=\{\alpha\in D:t_{\alpha}(s_{\alpha}-\zeta)\in L\setminus(\mathbf{V}_{\theta}+\mathbf{U}_{\theta})\}\in\mathcal{F}(I), H 0 = { α ∈ D : t α ( s α − ζ ) ∈ L ∖ ( V θ + U θ )} ∈ F ( I ) ,
H ζ 0 = { t α ( s α − ζ ) : α ∈ H 0 } ⊆ L ∖ ( V θ + U θ ) . H^{0}_{\zeta}=\{t_{\alpha}(s_{\alpha}-\zeta):\alpha\in H^{0}\}\subseteq L\setminus(\mathbf{V}_{\theta}+\mathbf{U}_{\theta}). H ζ 0 = { t α ( s α − ζ ) : α ∈ H 0 } ⊆ L ∖ ( V θ + U θ ) .
Therefore H ζ 0 ∩ ( V θ + U θ ) = ∅ H^{0}_{\zeta}\cap(\mathbf{V}_{\theta}+\mathbf{U}_{\theta})=\varnothing H ζ 0 ∩ ( V θ + U θ ) = ∅ , this shows that θ ∉ H ζ 0 \theta\notin H^{0}_{\zeta} θ ∈ / H ζ 0 which is a contradiction.
(ii) Let ζ \zeta ζ be a rough weighted I τ I_{\tau} I τ -cluster point of s = { s α } α ∈ D s=\{s_{\alpha}\}_{\alpha\in D} s = { s α } α ∈ D . Then for any τ \tau τ -neighbourhood U θ \mathbf{U}_{\theta} U θ of θ \theta θ we get
A = { α ∈ D : t α ( s α − ζ ) ∈ L ∖ ( V θ + U θ ) } ∉ F ( I ) . A=\{\alpha\in D:t_{\alpha}(s_{\alpha}-\zeta)\in L\setminus(\mathbf{V}_{\theta}+\mathbf{U}_{\theta})\}\notin\mathcal{F}(I). A = { α ∈ D : t α ( s α − ζ ) ∈ L ∖ ( V θ + U θ )} ∈ / F ( I ) .
If H ∈ F ( I ) H\in\mathcal{F}(I) H ∈ F ( I ) then H ⊈ A H\not\subseteq A H ⊆ A . Then there exists α 0 ∈ H \alpha_{0}\in H α 0 ∈ H such that α 0 ∈ D ∖ A \alpha_{0}\in D\setminus A α 0 ∈ D ∖ A (since α 0 ∉ A ) . T h e r e f o r e \alpha_{0}\notin A). Therefore α 0 ∈ / A ) . T h er e f or e t_{\alpha_{0}}(s_{\alpha_{0}}-\zeta)\in H_{\zeta}\cap(\mathbf{V}_{\theta}+\mathbf{U}_{\theta})$. Since L L L is Hausdorff and H ζ ∩ ( V θ + U θ ) ≠ ∅ H_{\zeta}\cap(\mathbf{V}_{\theta}+\mathbf{U}_{\theta})\ne\varnothing H ζ ∩ ( V θ + U θ ) = ∅ for all U θ \mathbf{U}_{\theta} U θ so H ζ ∩ V θ ≠ ∅ H_{\zeta}\cap\mathbf{V}_{\theta}\ne\varnothing H ζ ∩ V θ = ∅ . Hence our assertion follows.
We now derive a succinct characterisation of closed set in terms of the set R t V θ − I ( Γ s τ ) \mathcal{R}^{\mathbf{V}_{\theta}}_{t}-I(\Gamma^{\tau}_{s}) R t V θ − I ( Γ s τ ) .
Theorem 2.14. Let ( L , τ ) (L,\tau) ( L , τ ) be a completely separable locally solid Riesz space , U θ \mathbf{U}_{\theta} U θ be a closed τ \tau τ -neighbourhood of θ \theta θ , I I I be a non-trivial ideal of a directed set D D D , and there exist a pairwise disjoint sequence of sets { B i } i ∈ N \{B_{i}\}_{i\in\mathbb{N}} { B i } i ∈ N such that B i ⊂ D B_{i}\subset D B i ⊂ D , B i ∉ I B_{i}\notin I B i ∈ / I for all i ∈ N i\in\mathbb{N} i ∈ N . Then there is a net s = { s α } α ∈ D s=\{s_{\alpha}\}_{\alpha\in D} s = { s α } α ∈ D in L L L and a weighted net t = { t α } α ∈ D t=\{t_{\alpha}\}_{\alpha\in D} t = { t α } α ∈ D such that
U θ = ⋂ V θ ∈ N s o l ( R t V θ − I ( Γ s τ ) ) . \mathbf{U}_{\theta}=\bigcap_{\mathbf{V}_{\theta}\in\mathcal{N}_{\mathrm{sol}}}\left(\mathcal{R}^{\mathbf{V}_{\theta}}_{t}-I(\Gamma^{\tau}_{s})\right). U θ = V θ ∈ N sol ⋂ ( R t V θ − I ( Γ s τ ) ) .
Proof. Since U θ \mathbf{U}_{\theta} U θ is separable so there exists a countable dense set A = { a 1 , a 2 , … } A=\{a_{1},a_{2},\ldots\} A = { a 1 , a 2 , … } of U θ \mathbf{U}_{\theta} U θ such that A ‾ = U θ \overline{A}=\mathbf{U}_{\theta} A = U θ . Let s = { s α } α ∈ D s=\{s_{\alpha}\}_{\alpha\in D} s = { s α } α ∈ D be a net in the following manner:
s α = { a i if α ∈ B i , a 1 if α ∈ D ∖ ⋃ i = 1 ∞ B i . s_{\alpha}=
\begin{cases}
a_{i} & \text{if }\alpha\in B_{i},\\
a_{1} & \text{if }\alpha\in D\setminus\bigcup_{i=1}^{\infty}B_{i}.
\end{cases} s α = { a i a 1 if α ∈ B i , if α ∈ D ∖ ⋃ i = 1 ∞ B i .
We choose a I I I -bounded weighted net t = { t α } α ∈ D t=\{t_{\alpha}\}_{\alpha\in D} t = { t α } α ∈ D such that ∣ t α ∣ ≥ δ > 1 \lvert t_{\alpha}\rvert\ge\delta>1 ∣ t α ∣ ≥ δ > 1 for all α ∈ D \alpha\in D α ∈ D . From the I I I -boundedness of t = { t α } α ∈ D t=\{t_{\alpha}\}_{\alpha\in D} t = { t α } α ∈ D we get a large positive real number M M M such that
B = { α ∈ D : ∣ t α ∣ ≤ M } ∈ F ( I ) . B=\{\alpha\in D:\lvert t_{\alpha}\rvert\le M\}\in\mathcal{F}(I). B = { α ∈ D : ∣ t α ∣ ≤ M } ∈ F ( I ) .
Let u ∈ U θ u\in\mathbf{U}_{\theta} u ∈ U θ . Since u + 1 M V θ u+\frac{1}{M}\mathbf{V}_{\theta} u + M 1 V θ is a τ \tau τ -neighbourhood of u u u for all V θ ∈ N s o l \mathbf{V}_{\theta}\in\mathcal{N}_{\mathrm{sol}} V θ ∈ N sol . Then
A ∩ ( u + 1 M V θ ) ≠ ∅ ⟹ a i ∈ u + 1 M V θ for some i ∈ N ⟹ M ( a i − u ) ∈ V θ for some i ∈ N ⟹ t α ( s α − u ) ∈ V θ where α ∈ B ∩ B i for some i ∈ N ⟹ B ∩ B i ⊆ { α ∈ D : t α ( s α − u ) ∈ V θ + W θ } ∉ I , \begin{aligned}
A\cap\left(u+\frac{1}{M}\mathbf{V}_{\theta}\right)\ne\varnothing
&\Longrightarrow a_{i}\in u+\frac{1}{M}\mathbf{V}_{\theta}\text{ for some }i\in\mathbb{N}\\
&\Longrightarrow M(a_{i}-u)\in\mathbf{V}_{\theta}\text{ for some }i\in\mathbb{N}\\
&\Longrightarrow t_{\alpha}(s_{\alpha}-u)\in\mathbf{V}_{\theta}\text{ where }\alpha\in B\cap B_{i}\text{ for some }i\in\mathbb{N}\\
&\Longrightarrow B\cap B_{i}\subseteq\{\alpha\in D:t_{\alpha}(s_{\alpha}-u)\in\mathbf{V}_{\theta}+\mathbf{W}_{\theta}\}\notin I,
\end{aligned} A ∩ ( u + M 1 V θ ) = ∅ ⟹ a i ∈ u + M 1 V θ for some i ∈ N ⟹ M ( a i − u ) ∈ V θ for some i ∈ N ⟹ t α ( s α − u ) ∈ V θ where α ∈ B ∩ B i for some i ∈ N ⟹ B ∩ B i ⊆ { α ∈ D : t α ( s α − u ) ∈ V θ + W θ } ∈ / I ,
where W θ \mathbf{W}_{\theta} W θ is any τ \tau τ -neighbourhood of θ \theta θ (since B ∩ B i ∉ I B\cap B_{i}\notin I B ∩ B i ∈ / I , otherwise B i = ( N ∖ B ) ∩ ( B ∩ B i ) ∈ I B_{i}=(\mathbb{N}\setminus B)\cap(B\cap B_{i})\in I B i = ( N ∖ B ) ∩ ( B ∩ B i ) ∈ I , a contradiction). Therefore
u ∈ R t V θ − I ( Γ s τ ) ⟹ U θ ⊆ R t V θ − I ( Γ s τ ) , u\in\mathcal{R}^{\mathbf{V}_{\theta}}_{t}-I(\Gamma^{\tau}_{s})\Longrightarrow\mathbf{U}_{\theta}\subseteq\mathcal{R}^{\mathbf{V}_{\theta}}_{t}-I(\Gamma^{\tau}_{s}), u ∈ R t V θ − I ( Γ s τ ) ⟹ U θ ⊆ R t V θ − I ( Γ s τ ) ,
where V θ \mathbf{V}_{\theta} V θ is an arbitrary element of N s o l \mathcal{N}_{\mathrm{sol}} N sol . Then,
U θ ⊆ ⋂ V θ ∈ N s o l ( R t V θ − I ( Γ s τ ) ) . (8) \mathbf{U}_{\theta}\subseteq\bigcap_{\mathbf{V}_{\theta}\in\mathcal{N}_{\mathrm{sol}}}\left(\mathcal{R}^{\mathbf{V}_{\theta}}_{t}-I(\Gamma^{\tau}_{s})\right). \tag*{(8)} U θ ⊆ V θ ∈ N sol ⋂ ( R t V θ − I ( Γ s τ ) ) . (8)
Without using the finite intersection property or compactness, we derive the condition
⋂ V θ ∈ N s o l ( R t V θ − I ( Γ s τ ) ) ≠ ∅ \bigcap_{\mathbf{V}_{\theta}\in\mathcal{N}_{\mathrm{sol}}}\left(\mathcal{R}^{\mathbf{V}_{\theta}}_{t}-I(\Gamma^{\tau}_{s})\right)\ne\varnothing V θ ∈ N sol ⋂ ( R t V θ − I ( Γ s τ ) ) = ∅
from equations 2.3 and 2.4.
Next, we assume that
ζ ∈ ⋂ V θ ∈ N s o l ( R t V θ − I ( Γ s τ ) ) . \zeta\in\bigcap_{\mathbf{V}_{\theta}\in\mathcal{N}_{sol}}\left(\mathcal{R}_{t}^{\mathbf{V}_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau})\right). ζ ∈ V θ ∈ N so l ⋂ ( R t V θ − I ( Γ s τ ) ) .
Also let
C = ⋂ V θ ∈ N s o l { α ∈ D : t α ( s α − ζ ) ∈ V θ + V θ } . C=\bigcap_{\mathbf{V}_{\theta}\in\mathcal{N}_{sol}}\{\alpha\in D:t_{\alpha}(s_{\alpha}-\zeta)\in\mathbf{V}_{\theta}+\mathbf{V}_{\theta}\}. C = V θ ∈ N so l ⋂ { α ∈ D : t α ( s α − ζ ) ∈ V θ + V θ } .
If C = ∅ C=\varnothing C = ∅ , there exist two τ \tau τ -neighbourhoods V θ ′ , V θ ′ ′ ∈ N s o l \mathbf{V}_{\theta}^{\prime},\mathbf{V}_{\theta}^{\prime\prime}\in\mathcal{N}_{sol} V θ ′ , V θ ′′ ∈ N so l such that
{ α ∈ D : t α ( s α − ζ ) ∈ V θ ′ + V θ ′ } ∩ { α ∈ D : t α ( s α − ζ ) ∈ V θ ′ ′ + V θ ′ ′ } = ∅ . \{\alpha\in D:t_{\alpha}(s_{\alpha}-\zeta)\in\mathbf{V}_{\theta}^{\prime}+\mathbf{V}_{\theta}^{\prime}\}\cap\{\alpha\in D:t_{\alpha}(s_{\alpha}-\zeta)\in\mathbf{V}_{\theta}^{\prime\prime}+\mathbf{V}_{\theta}^{\prime\prime}\}=\varnothing. { α ∈ D : t α ( s α − ζ ) ∈ V θ ′ + V θ ′ } ∩ { α ∈ D : t α ( s α − ζ ) ∈ V θ ′′ + V θ ′′ } = ∅ .
Since 2 V θ ′ ∩ 2 V θ ′ ′ 2\mathbf{V}_{\theta}^{\prime}\cap2\mathbf{V}_{\theta}^{\prime\prime} 2 V θ ′ ∩ 2 V θ ′′ is a τ \tau τ -neighbourhood of θ \theta θ . Then there exists an element V θ ′ ′ ′ ∈ N s o l \mathbf{V}_{\theta}^{\prime\prime\prime}\in\mathcal{N}_{sol} V θ ′′′ ∈ N so l such that
V θ ′ ′ ′ + V θ ′ ′ ′ ⊆ 2 V θ ′ ∩ 2 V θ ′ ′ . \mathbf{V}_{\theta}^{\prime\prime\prime}+\mathbf{V}_{\theta}^{\prime\prime\prime}\subseteq2\mathbf{V}_{\theta}^{\prime}\cap2\mathbf{V}_{\theta}^{\prime\prime}. V θ ′′′ + V θ ′′′ ⊆ 2 V θ ′ ∩ 2 V θ ′′ .
Therefore { α ∈ D : t α ( s α − ζ ) ∈ V θ ′ ′ ′ + V θ ′ ′ ′ } = ∅ \{\alpha\in D:t_{\alpha}(s_{\alpha}-\zeta)\in\mathbf{V}_{\theta}^{\prime\prime\prime}+\mathbf{V}_{\theta}^{\prime\prime\prime}\}=\varnothing { α ∈ D : t α ( s α − ζ ) ∈ V θ ′′′ + V θ ′′′ } = ∅ , which is a contradiction since
ζ ∈ R t V θ ′ ′ ′ − I ( Γ s τ ) \zeta\in\mathcal{R}_{t}^{\mathbf{V}_{\theta}^{\prime\prime\prime}}-\mathcal{I}(\Gamma_{s}^{\tau}) ζ ∈ R t V θ ′′′ − I ( Γ s τ )
and { α ∈ D : t α ( s α − ζ ) ∈ V θ ′ ′ ′ + V θ ′ ′ ′ } ∉ I \{\alpha\in D:t_{\alpha}(s_{\alpha}-\zeta)\in\mathbf{V}_{\theta}^{\prime\prime\prime}+\mathbf{V}_{\theta}^{\prime\prime\prime}\}\notin\mathcal{I} { α ∈ D : t α ( s α − ζ ) ∈ V θ ′′′ + V θ ′′′ } ∈ / I .
Let α 0 ∈ C \alpha_{0}\in C α 0 ∈ C . Then t α 0 ( s α 0 − ζ ) ∈ V θ + V θ t_{\alpha_{0}}(s_{\alpha_{0}}-\zeta)\in\mathbf{V}_{\theta}+\mathbf{V}_{\theta} t α 0 ( s α 0 − ζ ) ∈ V θ + V θ for all V θ ∈ N s o l \mathbf{V}_{\theta}\in\mathcal{N}_{sol} V θ ∈ N so l .
⟹ δ ( a i − ζ ) ∈ V θ + V θ for some i ∈ N , and for all V θ ∈ N s o l \Longrightarrow\ \delta(a_i-\zeta)\in\mathbf{V}_{\theta}+\mathbf{V}_{\theta}\text{ for some }i\in\mathbb{N},\text{ and for all }\mathbf{V}_{\theta}\in\mathcal{N}_{sol} ⟹ δ ( a i − ζ ) ∈ V θ + V θ for some i ∈ N , and for all V θ ∈ N so l
⟹ a i ∈ ζ + 1 δ ( V θ + V θ ) ⊆ ζ + V θ + V θ for some i ∈ N , and for all V θ ∈ N s o l . \Longrightarrow\ a_i\in\zeta+\frac{1}{\delta}(\mathbf{V}_{\theta}+\mathbf{V}_{\theta})\subseteq\zeta+\mathbf{V}_{\theta}+\mathbf{V}_{\theta}\text{ for some }i\in\mathbb{N},\text{ and for all }\mathbf{V}_{\theta}\in\mathcal{N}_{sol}. ⟹ a i ∈ ζ + δ 1 ( V θ + V θ ) ⊆ ζ + V θ + V θ for some i ∈ N , and for all V θ ∈ N so l .
Therefore a i ∈ ( ζ + V θ + V θ ) ∩ U θ ≠ ∅ a_i\in(\zeta+\mathbf{V}_{\theta}+\mathbf{V}_{\theta})\cap U_{\theta}\ne\varnothing a i ∈ ( ζ + V θ + V θ ) ∩ U θ = ∅ for all V θ ∈ N s o l \mathbf{V}_{\theta}\in\mathcal{N}_{sol} V θ ∈ N so l . Since U θ U_{\theta} U θ is a closed set so ζ ∈ U θ \zeta\in U_{\theta} ζ ∈ U θ . □ \square □
In the year (2004/2005), Činčura et al. [7 ] give an important result related to the I \mathcal{I} I -cluster point set, which is stated below.
Theorem 2.15. [7 Lemma 3.1] Let I \mathcal{I} I be an admissible ideal on N \mathbb{N} N , Γ x ( I ) \Gamma_x(\mathcal{I}) Γ x ( I ) be a set of all I \mathcal{I} I -cluster points of a sequence x = { x n } n ∈ N x=\{x_n\}_{n\in\mathbb{N}} x = { x n } n ∈ N and K K K be a compact subset of a metric space X X X . Then for every sequence x = { x n } n ∈ N x=\{x_n\}_{n\in\mathbb{N}} x = { x n } n ∈ N in X X X with { n ∈ N : x n ∈ K } ∉ I \{n\in\mathbb{N}:x_n\in K\}\notin\mathcal{I} { n ∈ N : x n ∈ K } ∈ / I we have
K ∩ Γ x ( I ) ≠ ∅ . K\cap\Gamma_x(\mathcal{I})\ne\varnothing. K ∩ Γ x ( I ) = ∅ .
While reformulating the above results based on the set ‘rough weighted I τ \mathcal{I}_{\tau} I τ -cluster points of a net’ in locally solid Riesz spaces, the object of
K ∩ ( R t V θ − I ( Γ s τ ) ) ≠ ∅ K\cap\left(\mathcal{R}_{t}^{\mathbf{V}_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau})\right)\ne\varnothing K ∩ ( R t V θ − I ( Γ s τ ) ) = ∅
is violated. We exemplify this assertion below.
Example 2.16. Consider the locally solid Riesz space, directed set D D D , G G G and ideal I \mathcal{I} I as in Example 2.5 (i). Also let the compact set K = { x ∈ R 2 : ∥ x ∥ ≤ 1 } = { ( x , y ) ∈ R 2 : x 2 + y 2 ≤ 1 } = V θ K=\{x\in\mathbb{R}^{2}:\lVert x\rVert\le1\}=\{(x,y)\in\mathbb{R}^{2}:x^{2}+y^{2}\le1\}=\mathbf{V}_{\theta} K = { x ∈ R 2 : ∥ x ∥ ≤ 1 } = {( x , y ) ∈ R 2 : x 2 + y 2 ≤ 1 } = V θ , net s = { s α } α ∈ D s=\{s_{\alpha}\}_{\alpha\in D} s = { s α } α ∈ D and the weighted net t = { t α } α ∈ D t=\{t_{\alpha}\}_{\alpha\in D} t = { t α } α ∈ D are in the following manner:
s U θ ( r ) = { ( r , 0 ) if U θ ( r ) ∈ G , ( 1 r , 1 r ) if U θ ( r ) ∈ D ∖ G , and t U θ ( r ) = { − ( r + 1 ) if r ∈ G , 2 r + 1 if r ∈ D ∖ G . s_{U_{\theta}(r)}=
\begin{cases}
(r,0)&\text{if }U_{\theta}(r)\in G,\\
\left(\frac{1}{r},\frac{1}{r}\right)&\text{if }U_{\theta}(r)\in D\setminus G,
\end{cases}
\qquad\text{and}\qquad
t_{U_{\theta}(r)}=
\begin{cases}
-(r+1)&\text{if }r\in G,\\
2r+1&\text{if }r\in D\setminus G.
\end{cases} s U θ ( r ) = { ( r , 0 ) ( r 1 , r 1 ) if U θ ( r ) ∈ G , if U θ ( r ) ∈ D ∖ G , and t U θ ( r ) = { − ( r + 1 ) 2 r + 1 if r ∈ G , if r ∈ D ∖ G . Therefore K ∩ { R t V θ − I ( Γ s τ ) } = ∅ K\cap\{\mathcal{R}_{t}^{\mathbf{V}_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau})\}=\varnothing K ∩ { R t V θ − I ( Γ s τ )} = ∅ and { α ∈ D : s α ∈ K } ∉ I \{\alpha\in D:s_{\alpha}\in K\}\notin\mathcal{I} { α ∈ D : s α ∈ K } ∈ / I .
Theorem 2.17. Let I \mathcal{I} I be a non-trivial ideal of a directed set D D D , ( L , τ ) (L,\tau) ( L , τ ) be a locally solid Riesz space, V θ ∈ N s o l \mathbf{V}_{\theta}\in\mathcal{N}_{sol} V θ ∈ N so l , K K K be a compact subset of L L L and the weighted net t = { t α } α ∈ D t=\{t_{\alpha}\}_{\alpha\in D} t = { t α } α ∈ D be I \mathcal{I} I -bounded. Then for every net s = { s α } α ∈ D s=\{s_{\alpha}\}_{\alpha\in D} s = { s α } α ∈ D in L L L with
{ α ∈ D : s α ∈ K } ∉ I \{\alpha\in D:s_{\alpha}\in K\}\notin\mathcal{I} { α ∈ D : s α ∈ K } ∈ / I we have K ∩ { R t V θ − I ( Γ s τ ) } ≠ ∅ K\cap\{\mathcal{R}_{t}^{\mathbf{V}_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau})\}\ne\varnothing K ∩ { R t V θ − I ( Γ s τ )} = ∅ .
Proof. We assume by way of contradiction, let K ∩ { R t V θ − I ( Γ s τ ) } = ∅ K \cap\{\mathcal{R}_{t}^{V_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau})\}=\varnothing K ∩ { R t V θ − I ( Γ s τ )} = ∅ . Let k ∈ K k \in K k ∈ K . Then k ∉ R t V θ − I ( Γ s τ ) k \notin\mathcal{R}_{t}^{V_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau}) k ∈ / R t V θ − I ( Γ s τ ) , there exists a U θ ∈ N s o l U_{\theta} \in\mathcal{N}_{\mathrm{sol}} U θ ∈ N sol such that
A k = { α ∈ D : t α ( s α − k ) ∈ V θ + U θ } ∈ I . A_k=\{\alpha\in D:t_{\alpha}(s_{\alpha}-k)\in V_{\theta}+U_{\theta}\}\in\mathcal{I}. A k = { α ∈ D : t α ( s α − k ) ∈ V θ + U θ } ∈ I .
From the I \mathcal{I} I -boundedness of the weighted net t = { t α } α ∈ D t=\{t_{\alpha}\}_{\alpha\in D} t = { t α } α ∈ D there exists M > 0 M>0 M > 0 such that
B = { α ∈ D : ∣ t α ∣ ≤ M } ∈ F ( I ) . B=\{\alpha\in D:|t_{\alpha}|\leq M\}\in\mathcal{F}(\mathcal{I}). B = { α ∈ D : ∣ t α ∣ ≤ M } ∈ F ( I ) .
Also
K ⊆ ⋃ k ∈ K ( k + V θ M ) . K\subseteq\bigcup_{k\in K}\left(k+\frac{V_{\theta}}{M}\right). K ⊆ k ∈ K ⋃ ( k + M V θ ) .
Since K K K is a compact set, then there exists a finite number of elements k 1 , k 2 , … , k n k_1,k_2,\ldots,k_n k 1 , k 2 , … , k n in K K K such that
K ⊆ ⋃ i = 1 n ( k i + V θ M ) . K\subseteq\bigcup_{i=1}^{n}\left(k_i+\frac{V_{\theta}}{M}\right). K ⊆ i = 1 ⋃ n ( k i + M V θ ) .
Let C = { α ∈ D : s α ∈ K } ∉ I C=\{\alpha\in D:s_{\alpha}\in K\}\notin\mathcal{I} C = { α ∈ D : s α ∈ K } ∈ / I . Then B ∩ C ≠ ∅ B\cap C\neq\varnothing B ∩ C = ∅ (otherwise C ⊆ D ∖ B ∈ I C\subseteq D\setminus B\in\mathcal{I} C ⊆ D ∖ B ∈ I , a contradiction). If α ∈ B ∩ C \alpha\in B\cap C α ∈ B ∩ C then
s α ∈ K ⊆ ⋃ i = 1 n ( k i + V θ M ) ⟹ s α ∈ k i + V θ M for some i ∈ { 1 , 2 , 3 , … , n } , s_{\alpha}\in K\subseteq\bigcup_{i=1}^{n}\left(k_i+\frac{V_{\theta}}{M}\right)\Longrightarrow s_{\alpha}\in k_i+\frac{V_{\theta}}{M}\text{ for some }i\in\{1,2,3,\ldots,n\}, s α ∈ K ⊆ i = 1 ⋃ n ( k i + M V θ ) ⟹ s α ∈ k i + M V θ for some i ∈ { 1 , 2 , 3 , … , n } ,
⟹ M ( s α − k i ) ∈ V θ for some i ∈ { 1 , 2 , 3 , … , n } , \Longrightarrow M(s_{\alpha}-k_i)\in V_{\theta}\text{ for some }i\in\{1,2,3,\ldots,n\}, ⟹ M ( s α − k i ) ∈ V θ for some i ∈ { 1 , 2 , 3 , … , n } ,
⟹ t α ( s α − k i ) ∈ V θ for some i ∈ { 1 , 2 , 3 , … , n } , \Longrightarrow t_{\alpha}(s_{\alpha}-k_i)\in V_{\theta}\text{ for some }i\in\{1,2,3,\ldots,n\}, ⟹ t α ( s α − k i ) ∈ V θ for some i ∈ { 1 , 2 , 3 , … , n } ,
since ∣ t α ( s α − k i ) ∣ ≤ ∣ M ( s α − k i ) ∣ |t_{\alpha}(s_{\alpha}-k_i)|\leq|M(s_{\alpha}-k_i)| ∣ t α ( s α − k i ) ∣ ≤ ∣ M ( s α − k i ) ∣ for all i ∈ { 1 , 2 , 3 , … , n } i\in\{1,2,3,\ldots,n\} i ∈ { 1 , 2 , 3 , … , n } , α ∈ B ∩ C \alpha\in B\cap C α ∈ B ∩ C and V θ ∈ N s o l V_{\theta}\in\mathcal{N}_{\mathrm{sol}} V θ ∈ N sol . Therefore
B ∩ C ⊆ { α ∈ D : t α ( s α − k i ) ∈ V θ } for some i ∈ { 1 , 2 , 3 , … , n } , B\cap C\subseteq\{\alpha\in D:t_{\alpha}(s_{\alpha}-k_i)\in V_{\theta}\}\text{ for some }i\in\{1,2,3,\ldots,n\}, B ∩ C ⊆ { α ∈ D : t α ( s α − k i ) ∈ V θ } for some i ∈ { 1 , 2 , 3 , … , n } ,
⟹ B ∩ C ⊆ ⋃ i = 1 n { α ∈ D : t α ( s α − k i ) ∈ V θ } ⊆ ⋃ i = 1 n { α ∈ D : t α ( s α − k i ) ∈ V θ + U θ } = ⋃ i = 1 n A k i ∈ I . \Longrightarrow B\cap C\subseteq\bigcup_{i=1}^{n}\{\alpha\in D:t_{\alpha}(s_{\alpha}-k_i)\in V_{\theta}\}\subseteq\bigcup_{i=1}^{n}\{\alpha\in D:t_{\alpha}(s_{\alpha}-k_i)\in V_{\theta}+U_{\theta}\}=\bigcup_{i=1}^{n}A_{k_i}\in\mathcal{I}. ⟹ B ∩ C ⊆ i = 1 ⋃ n { α ∈ D : t α ( s α − k i ) ∈ V θ } ⊆ i = 1 ⋃ n { α ∈ D : t α ( s α − k i ) ∈ V θ + U θ } = i = 1 ⋃ n A k i ∈ I .
Finally, C ⊆ ( D ∖ B ) ∪ ( B ∩ C ) ∈ I C\subseteq(D\setminus B)\cup(B\cap C)\in\mathcal{I} C ⊆ ( D ∖ B ) ∪ ( B ∩ C ) ∈ I , a contradiction. Therefore, this is the result. □
Theorem 2.18
Theorem 2.18. Let I \mathcal{I} I be a non-trivial ideal of a directed set D D D , t = { t α } α ∈ D t=\{t_{\alpha}\}_{\alpha\in D} t = { t α } α ∈ D be a weighted net, s = { s α } α ∈ D s=\{s_{\alpha}\}_{\alpha\in D} s = { s α } α ∈ D be a net in a locally solid Riesz space L L L , V θ ∈ N s o l V_{\theta}\in\mathcal{N}_{\mathrm{sol}} V θ ∈ N sol and V θ V_{\theta} V θ be a compact τ \tau τ -neighbourhood of θ \theta θ . If R t V θ − I ( Γ s τ ) ≠ ∅ \mathcal{R}_{t}^{V_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau})\neq\varnothing R t V θ − I ( Γ s τ ) = ∅ then there exists a compact set K K K of L L L such that
{ α ∈ D : s α ∈ K } ∉ I . \{\alpha\in D:s_{\alpha}\in K\}\notin\mathcal{I}. { α ∈ D : s α ∈ K } ∈ / I .
Proof. Let ζ ∈ R t V θ − I ( Γ s τ ) \zeta\in\mathcal{R}_{t}^{V_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau}) ζ ∈ R t V θ − I ( Γ s τ ) . Then for every U θ ∈ N s o l U_{\theta}\in\mathcal{N}_{\mathrm{sol}} U θ ∈ N sol and U θ ⊆ V θ U_{\theta}\subseteq V_{\theta} U θ ⊆ V θ such that
{ α ∈ D : t α ( s α − ζ ) ∈ V θ + U θ } ∉ I . \{\alpha\in D:t_{\alpha}(s_{\alpha}-\zeta)\in V_{\theta}+U_{\theta}\}\notin\mathcal{I}. { α ∈ D : t α ( s α − ζ ) ∈ V θ + U θ } ∈ / I .
⟹ A = { α ∈ D : t α ( s α − ζ ) ∈ V θ + V θ } ∉ I . \Longrightarrow A=\{\alpha\in D:t_{\alpha}(s_{\alpha}-\zeta)\in V_{\theta}+V_{\theta}\}\notin\mathcal{I}. ⟹ A = { α ∈ D : t α ( s α − ζ ) ∈ V θ + V θ } ∈ / I .
Since t = { t α } α ∈ D t=\{t_{\alpha}\}_{\alpha\in D} t = { t α } α ∈ D is a weighted net, there exists a positive real number δ \delta δ such that ∣ t α ∣ ≥ δ |t_{\alpha}|\geq\delta ∣ t α ∣ ≥ δ for all α ∈ D \alpha\in D α ∈ D . Let α ∈ A \alpha\in A α ∈ A . Then ∣ δ ( s α − ζ ) ∣ ≤ ∣ t α ( s α − ζ ) ∣ |\delta(s_{\alpha}-\zeta)|\leq|t_{\alpha}(s_{\alpha}-\zeta)| ∣ δ ( s α − ζ ) ∣ ≤ ∣ t α ( s α − ζ ) ∣ for all α ∈ A \alpha\in A α ∈ A . Also, V θ + V θ ∈ N s o l V_{\theta}+V_{\theta}\in\mathcal{N}_{\mathrm{sol}} V θ + V θ ∈ N sol and t α ( s α − ζ ) ∈ V θ + V θ t_{\alpha}(s_{\alpha}-\zeta)\in V_{\theta}+V_{\theta} t α ( s α − ζ ) ∈ V θ + V θ implies δ ( s α − ζ ) ∈ V θ + V θ \delta(s_{\alpha}-\zeta)\in V_{\theta}+V_{\theta} δ ( s α − ζ ) ∈ V θ + V θ . Therefore
A ⊆ { α ∈ D : δ ( s α − ζ ) ∈ V θ + V θ } = { α ∈ D : s α ∈ K } ∉ I A\subseteq\{\alpha\in D:\delta(s_{\alpha}-\zeta)\in V_{\theta}+V_{\theta}\}=\{\alpha\in D:s_{\alpha}\in K\}\notin\mathcal{I} A ⊆ { α ∈ D : δ ( s α − ζ ) ∈ V θ + V θ } = { α ∈ D : s α ∈ K } ∈ / I
where K = ζ + 1 δ ( V θ + V θ ) K=\zeta+\frac{1}{\delta}(V_{\theta}+V_{\theta}) K = ζ + δ 1 ( V θ + V θ ) and K K K is a compact set in L L L . □
The following example shows that if V θ V_{\theta} V θ is not a compact τ \tau τ -neighbourhood of θ \theta θ in the above theorem, then there does not exists any compact set K K K of L L L such that { α ∈ D : s α ∈ K } ∉ I \{\alpha\in D:s_{\alpha}\in K\}\notin\mathcal{I} { α ∈ D : s α ∈ K } ∈ / I .
Example 2.19. Let us consider the infinite dimensional norm linear space l 2 l^2 l 2 space with the norm
∥ x ∥ = ( ∑ i = 1 ∞ ∣ x i ∣ 2 ) \|x\|=\sqrt{\left(\sum_{i=1}^{\infty}|x_i|^2\right)} ∥ x ∥ = ( i = 1 ∑ ∞ ∣ x i ∣ 2 ) where x = { x n } n ∈ N ∈ l 2 x=\{x_n\}_{n\in\mathbb{N}}\in l^2 x = { x n } n ∈ N ∈ l 2 with coordinatewise ordering. Since l 2 l^2 l 2 is a locally solid Riesz space . The family N s o l \mathcal{N}_{sol} N so l of all U θ ( r ) U_\theta(r) U θ ( r ) defined as U θ ( r ) = { x ∈ l 2 : ∥ x ∥ < r } U_\theta(r)=\{x\in l^2:\|x\|<r\} U θ ( r ) = { x ∈ l 2 : ∥ x ∥ < r } where r > 0 r>0 r > 0 , constitutes a base at θ = ( 0 , 0 , … ) \theta=(0,0,\ldots) θ = ( 0 , 0 , … ) . The set D = N s o l D=\mathcal{N}_{sol} D = N so l is a directed set with the binary relation ≥ \geq ≥ defined by U θ ≥ V θ U_\theta\geq V_\theta U θ ≥ V θ if U θ ⊇ V θ U_\theta\supseteq V_\theta U θ ⊇ V θ for any U θ , V θ ∈ D U_\theta,V_\theta\in D U θ , V θ ∈ D . Consider the following decomposition of R + \mathbb{R}^{+} R + ,
B 0 = ( 0 , 1 ) ∪ { ⋃ n ∈ O [ n , n + 1 ] } where O = { 1 , 3 , 5 , … } , B_0=(0,1)\cup\left\{\bigcup_{n\in O}[n,n+1]\right\}\text{ where }O=\{1,3,5,\ldots\}, B 0 = ( 0 , 1 ) ∪ { n ∈ O ⋃ [ n , n + 1 ] } where O = { 1 , 3 , 5 , … } , B ℓ = ⋃ n ∈ O ( 2 ℓ n , 2 ℓ n + 1 ) where ℓ ∈ N = { 1 , 2 , 3 , … } , B_\ell=\bigcup_{n\in O}(2^\ell n,2^\ell n+1)\text{ where }\ell\in\mathbb{N}=\{1,2,3,\ldots\}, B ℓ = n ∈ O ⋃ ( 2 ℓ n , 2 ℓ n + 1 ) where ℓ ∈ N = { 1 , 2 , 3 , … } , ⟹ R + = ⋃ ℓ ∈ N 0 B ℓ where B i ∩ B j = ∅ for all i , j ∈ N 0 = { 0 , 1 , 2 , 3 , … } and i ≠ j . \Longrightarrow\mathbb{R}^{+}=\bigcup_{\ell\in\mathbb{N}_0}B_\ell\text{ where }B_i\cap B_j=\varnothing\text{ for all }i,j\in\mathbb{N}_0=\{0,1,2,3,\ldots\}\text{ and }i\ne j. ⟹ R + = ℓ ∈ N 0 ⋃ B ℓ where B i ∩ B j = ∅ for all i , j ∈ N 0 = { 0 , 1 , 2 , 3 , … } and i = j . Let I = { { U θ ( r ) : r ∈ Δ ( ⊂ R ) } : Δ can intersect finitely many elements of B i , i ∈ N 0 } \mathcal{I}=\{\{U_\theta(r):r\in\Delta(\subset\mathbb{R})\}:\Delta\text{ can intersect finitely many elements of }B_i,\ i\in\mathbb{N}_0\} I = {{ U θ ( r ) : r ∈ Δ ( ⊂ R )} : Δ can intersect finitely many elements of B i , i ∈ N 0 } . We choose the net s = { s α } α ∈ D s=\{s_\alpha\}_{\alpha\in D} s = { s α } α ∈ D and the weighted net t = { t α } α ∈ D t=\{t_\alpha\}_{\alpha\in D} t = { t α } α ∈ D are respectively:
s U θ ( r ) = { θ if r ∈ B 0 , e n if r ∈ B n , and t U ( r ) = { r + 1 if r ∈ B 0 , 1 if r ∈ B n , s_{U_\theta(r)}=
\begin{cases}
\theta& \text{if }r\in B_0,\\
e_n & \text{if }r\in B_n,
\end{cases}
\qquad\text{and}\qquad
t_{U(r)}=
\begin{cases}
r+1 & \text{if }r\in B_0,\\
1 & \text{if }r\in B_n,
\end{cases} s U θ ( r ) = { θ e n if r ∈ B 0 , if r ∈ B n , and t U ( r ) = { r + 1 1 if r ∈ B 0 , if r ∈ B n , where e n = ( δ j n ) e_n=(\delta_j^n) e n = ( δ j n ) has n th n^{\text{th}} n th term 1 and all other terms are 0. Let V θ = { x ∈ l 2 : ∥ x ∥ ≤ 1 } V_\theta=\{x\in l^2:\|x\|\leq1\} V θ = { x ∈ l 2 : ∥ x ∥ ≤ 1 } . For any τ \tau τ -neighbourhood U θ U_\theta U θ of θ \theta θ the set { α ∈ D : t α ( s α − θ ) ∈ V θ + U θ } = D ∉ I \{\alpha\in D:t_\alpha(s_\alpha-\theta)\in V_\theta+U_\theta\}=D\notin\mathcal{I} { α ∈ D : t α ( s α − θ ) ∈ V θ + U θ } = D ∈ / I . Therefore θ ∈ R t V θ − I ( Γ s τ ) ≠ ∅ \theta\in\mathcal{R}_{t}^{V_\theta}-\mathcal{I}(\Gamma_s^\tau)\ne\varnothing θ ∈ R t V θ − I ( Γ s τ ) = ∅ .
Next, let K K K be any subset of L L L . Then two possibility are their
(i): Let K ∩ ( { e i : i ∈ N } ∪ { θ } ) = { e m 1 , e m 2 , e m 3 , … , e m k } K\cap(\{e_i:i\in\mathbb{N}\}\cup\{\theta\})=\{e_{m_1},e_{m_2},e_{m_3},\ldots,e_{m_k}\} K ∩ ({ e i : i ∈ N } ∪ { θ }) = { e m 1 , e m 2 , e m 3 , … , e m k } be a finite set. Then
{ α ∈ D : s α ∈ K } = { U θ ( r ) : r ∈ ⋃ i = 1 k B m i } ∈ I . \{\alpha\in D:s_\alpha\in K\}=\{U_\theta(r):r\in\bigcup_{i=1}^{k}B_{m_i}\}\in\mathcal{I}. { α ∈ D : s α ∈ K } = { U θ ( r ) : r ∈ i = 1 ⋃ k B m i } ∈ I . (ii): Let K ∩ ( { e i : i ∈ N } ∪ { θ } ) = { e m 1 , e m 2 , e m 3 , … } K\cap(\{e_i:i\in\mathbb{N}\}\cup\{\theta\})=\{e_{m_1},e_{m_2},e_{m_3},\ldots\} K ∩ ({ e i : i ∈ N } ∪ { θ }) = { e m 1 , e m 2 , e m 3 , … } be an infinite set. Then the set
{ α ∈ D : s α ∈ K } = { U θ ( r ) : r ∈ ⋃ i = 1 ∞ B m i } ∉ I . \{\alpha\in D:s_\alpha\in K\}=\{U_\theta(r):r\in\bigcup_{i=1}^{\infty}B_{m_i}\}\notin\mathcal{I}. { α ∈ D : s α ∈ K } = { U θ ( r ) : r ∈ i = 1 ⋃ ∞ B m i } ∈ / I . Therefore the only possibility the set { α ∈ D : s α ∈ K } ∉ I \{\alpha\in D:s_\alpha\in K\}\notin\mathcal{I} { α ∈ D : s α ∈ K } ∈ / I if K K K contains infinitely many e i e_i e i where i ∈ N i\in\mathbb{N} i ∈ N . But this is not possible for any compact set K K K of l 2 l^2 l 2 .
In general if V θ V_\theta V θ be τ \tau τ -bounded (or τ \tau τ -unbounded) then the set R t V θ − I ( Γ s τ ) \mathcal{R}_{t}^{V_\theta}-\mathcal{I}(\Gamma_s^\tau) R t V θ − I ( Γ s τ ) may be τ \tau τ -unbounded. To prove this assertion, we consider the following examples:
Example 2.20. Case (i): Let V θ V_\theta V θ be τ \tau τ -bounded. To show that the set R t V θ − I ( Γ s τ ) \mathcal{R}_{t}^{V_\theta}-\mathcal{I}(\Gamma_s^\tau) R t V θ − I ( Γ s τ ) is τ \tau τ -unbounded. This example based on Example 2.9 (i). Replacing the net s = { s α } α ∈ D s=\{s_\alpha\}_{\alpha\in D} s = { s α } α ∈ D by the net s ∗ = { s α ∗ } α ∈ D s^*=\{s^*_\alpha\}_{\alpha\in D} s ∗ = { s α ∗ } α ∈ D where
s U θ ( r ) ∗ = { ( 1 , 1 ) if r ∈ B 0 , ( ℓ + 1 r 2 , 1 ) if r ∈ B ℓ , where ℓ ∈ N , s^*_{U_\theta(r)}=
\begin{cases}
(1,1) & \text{if }r\in B_0,\\
\left(\ell+\frac{1}{r^2},1\right) & \text{if }r\in B_\ell,\text{ where }\ell\in\mathbb{N},
\end{cases} s U θ ( r ) ∗ = { ( 1 , 1 ) ( ℓ + r 2 1 , 1 ) if r ∈ B 0 , if r ∈ B ℓ , where ℓ ∈ N , and keeping all other conditions unaltered in Example 2.9, consequently we get
N × { 1 } ⊆ R t V θ − I ( Γ s ∗ τ ) . \mathbb{N}\times\{1\}\subseteq\mathcal{R}_{t}^{V_\theta}-\mathcal{I}(\Gamma_{s^*}^\tau). N × { 1 } ⊆ R t V θ − I ( Γ s ∗ τ ) .
Case (ii): Let V θ \mathbf{V}_{\theta} V θ be τ \tau τ -unbounded. Consider the locally solid Riesz space , directed set D D D , a subset G G G of D D D and ideal I \mathcal{I} I as in Example 2.5 (i). Let the net s = { s α } α ∈ D s=\{s_{\alpha}\}_{\alpha\in D} s = { s α } α ∈ D and the weighted net t = { t α } α ∈ D t=\{t_{\alpha}\}_{\alpha\in D} t = { t α } α ∈ D be in the following manner:
s U θ ( r ) ∈ { ( r , 0 ) if U θ ( r ) ∈ G , ( 4 , 4 ) + ( 4 cos r , 4 sin r ) if U θ ( r ) ∈ D ∖ G , and t U θ ( r ) = { − ( r + 1 ) if r ∈ G , r + 1 if r ∈ D ∖ G . s_{\mathbf{U}_{\theta}(r)} \in\begin{cases} (r,0) & \text{if } \mathbf{U}_{\theta}(r)\in G, \\ (4,4)+(4\cos r,4\sin r) & \text{if } \mathbf{U}_{\theta}(r)\in D\setminus G, \end{cases}
\qquad\text{and}\qquad
t_{\mathbf{U}_{\theta}(r)}=\begin{cases} -(r+1) & \text{if } r\in G, \\ r+1 & \text{if } r\in D\setminus G. \end{cases} s U θ ( r ) ∈ { ( r , 0 ) ( 4 , 4 ) + ( 4 cos r , 4 sin r ) if U θ ( r ) ∈ G , if U θ ( r ) ∈ D ∖ G , and t U θ ( r ) = { − ( r + 1 ) r + 1 if r ∈ G , if r ∈ D ∖ G .
Let V θ = { ( x , y ) ∈ R 2 : − 1 < y < 1 } \mathbf{V}_{\theta}=\{(x,y)\in\mathbb{R}^{2}:-1<y<1\} V θ = {( x , y ) ∈ R 2 : − 1 < y < 1 } . Then R t V θ − I ( Γ s τ ) = { ( x , y ) ∈ R 2 : y = 0 } \mathcal{R}_{t}^{\mathbf{V}_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau})=\{(x,y)\in\mathbb{R}^{2}:y=0\} R t V θ − I ( Γ s τ ) = {( x , y ) ∈ R 2 : y = 0 } . This example also indicates the net s = { s α } α ∈ D s=\{s_{\alpha}\}_{\alpha\in D} s = { s α } α ∈ D is not rough weighted I τ \mathcal{I}_{\tau} I τ -convergent to any point of R 2 \mathbb{R}^{2} R 2 w.r.t the roughness of degree V θ \mathbf{V}_{\theta} V θ yet the set R t V θ − I ( Γ s τ ) \mathcal{R}_{t}^{\mathbf{V}_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau}) R t V θ − I ( Γ s τ ) is uncountable and τ \tau τ -unbounded.
We give some elementary properties concerning τ \tau τ -boundedness of the set R t V θ − I ( Γ s τ ) \mathcal{R}_{t}^{\mathbf{V}_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau}) R t V θ − I ( Γ s τ ) .
Proposition 2.21. Let ( L , τ ) (L,\tau) ( L , τ ) be a locally solid Riesz space, s = { s α } α ∈ D s=\{s_{\alpha}\}_{\alpha\in D} s = { s α } α ∈ D be weighted I τ \mathcal{I}_{\tau} I τ -bounded, t = { t α } α ∈ D t=\{t_{\alpha}\}_{\alpha\in D} t = { t α } α ∈ D be a weighted net, T = inf α ∈ D ∣ t α ∣ T=\inf_{\alpha\in D}|t_{\alpha}| T = inf α ∈ D ∣ t α ∣ and V θ ∈ N s o l \mathbf{V}_{\theta}\in\mathcal{N}_{sol} V θ ∈ N so l . Then
(i): R t V θ − I ( Γ s τ ) ⊆ 2 + μ T V θ \mathcal{R}_{t}^{\mathbf{V}_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau})\subseteq\frac{2+\mu}{T}\mathbf{V}_{\theta} R t V θ − I ( Γ s τ ) ⊆ T 2 + μ V θ , where μ \mu μ is a positive real number,
(ii): R t V θ − I ( Γ s τ ) \mathcal{R}_{t}^{\mathbf{V}_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau}) R t V θ − I ( Γ s τ ) is τ \tau τ -bounded if V θ \mathbf{V}_{\theta} V θ is τ \tau τ -bounded.
Proof. Since s = { s α } α ∈ D s=\{s_{\alpha}\}_{\alpha\in D} s = { s α } α ∈ D is weighted I τ \mathcal{I}_{\tau} I τ -bounded, then for every τ \tau τ -neighbourhood V θ \mathbf{V}_{\theta} V θ of θ \theta θ there exists some λ > 0 \lambda>0 λ > 0 such that A = { α ∈ D : λ t α s α ∉ V θ } ∈ I A=\{\alpha\in D:\lambda t_{\alpha}s_{\alpha}\notin\mathbf{V}_{\theta}\}\in\mathcal{I} A = { α ∈ D : λ t α s α ∈ / V θ } ∈ I . If ζ ∈ R t V θ − I ( Γ s τ ) \zeta\in\mathcal{R}_{t}^{\mathbf{V}_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau}) ζ ∈ R t V θ − I ( Γ s τ ) , then B = { α ∈ D : t α ( s α − ζ ) ∈ V θ + V θ } ∉ I B=\{\alpha\in D:t_{\alpha}(s_{\alpha}-\zeta)\in\mathbf{V}_{\theta}+\mathbf{V}_{\theta}\}\notin\mathcal{I} B = { α ∈ D : t α ( s α − ζ ) ∈ V θ + V θ } ∈ / I . Also, ( D ∖ A ) ∩ B ≠ ∅ (D\setminus A)\cap B\ne\varnothing ( D ∖ A ) ∩ B = ∅ (otherwise B ⊆ A ∈ I B\subseteq A\in\mathcal{I} B ⊆ A ∈ I , a contradiction). If α ∈ ( D ∖ A ) ∩ B \alpha\in(D\setminus A)\cap B α ∈ ( D ∖ A ) ∩ B , then
t α ζ = t α ( ζ − s α ) + t α s α ∈ ( 2 + 1 λ ) V θ . t_{\alpha}\zeta=t_{\alpha}(\zeta-s_{\alpha})+t_{\alpha}s_{\alpha}\in\left(2+\frac{1}{\lambda}\right)\mathbf{V}_{\theta}. t α ζ = t α ( ζ − s α ) + t α s α ∈ ( 2 + λ 1 ) V θ .
Since V θ ∈ N s o l \mathbf{V}_{\theta}\in\mathcal{N}_{sol} V θ ∈ N so l then T ζ ∈ ( 2 + 1 λ ) V θ T\zeta\in\left(2+\frac{1}{\lambda}\right)\mathbf{V}_{\theta} T ζ ∈ ( 2 + λ 1 ) V θ . Hence R t V θ − I ( Γ s τ ) ⊆ 2 + μ T V θ \mathcal{R}_{t}^{\mathbf{V}_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau})\subseteq\frac{2+\mu}{T}\mathbf{V}_{\theta} R t V θ − I ( Γ s τ ) ⊆ T 2 + μ V θ where μ = 1 λ \mu=\frac{1}{\lambda} μ = λ 1 .
For the next part V θ \mathbf{V}_{\theta} V θ is τ \tau τ -bounded so 2 + μ T V θ \frac{2+\mu}{T}\mathbf{V}_{\theta} T 2 + μ V θ also. □
As an application of the sets W I τ − L I M t V θ s \mathrm{WI}_{\tau}-\mathrm{LIM}_{t}^{\mathbf{V}_{\theta}}s WI τ − LIM t V θ s and R t V θ − I ( Γ s τ ) \mathcal{R}_{t}^{\mathbf{V}_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau}) R t V θ − I ( Γ s τ ) we prove the following result which characterise the ideal.
Theorem 2.22. Let ( L , τ ) (L,\tau) ( L , τ ) be a Hausdorff locally solid Riesz space , I \mathcal{I} I be a non-trivial ideal of a directed set D D D . Then I \mathcal{I} I is a maximal ideal on D D D if and only if W I τ − L I M t V θ s = R t V θ − I ( Γ s τ ) \mathrm{WI}_{\tau}-\mathrm{LIM}_{t}^{\mathbf{V}_{\theta}}s=\mathcal{R}_{t}^{\mathbf{V}_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau}) WI τ − LIM t V θ s = R t V θ − I ( Γ s τ ) for any net s = { s α } α ∈ D s=\{s_{\alpha}\}_{\alpha\in D} s = { s α } α ∈ D in L L L , with some roughness degree V θ \mathbf{V}_{\theta} V θ .
Proof. Let I \mathcal{I} I be a maximal ideal on D D D . From Definition 2.2 and 2.3 we get
W I τ − L I M t V θ s ⊆ R t V θ − I ( Γ s τ ) \mathrm{WI}_{\tau}-\mathrm{LIM}_{t}^{\mathbf{V}_{\theta}}s\subseteq\mathcal{R}_{t}^{\mathbf{V}_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau}) WI τ − LIM t V θ s ⊆ R t V θ − I ( Γ s τ )
for any roughness degree V θ ∈ N s o l \mathbf{V}_{\theta}\in\mathcal{N}_{sol} V θ ∈ N so l .
Next let, ζ ∈ R t V θ − I ( Γ s τ ) \zeta\in\mathcal{R}_{t}^{\mathbf{V}_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau}) ζ ∈ R t V θ − I ( Γ s τ ) . Then A = { α ∈ D : t α ( s α − ζ ) ∈ V θ + U θ } ∉ I A=\{\alpha\in D:t_{\alpha}(s_{\alpha}-\zeta)\in\mathbf{V}_{\theta}+\mathbf{U}_{\theta}\}\notin\mathcal{I} A = { α ∈ D : t α ( s α − ζ ) ∈ V θ + U θ } ∈ / I , for any τ \tau τ -neighbourhood U θ \mathbf{U}_{\theta} U θ of θ \theta θ . By the property of maximal ideal we get D ∖ A ∈ I D\setminus A\in\mathcal{I} D ∖ A ∈ I . This implies
D ∖ A = { α ∈ D : t α ( s α − ζ ) ∉ V θ + U θ } ∈ I . D\setminus A=\{\alpha\in D:t_{\alpha}(s_{\alpha}-\zeta)\notin\mathbf{V}_{\theta}+\mathbf{U}_{\theta}\}\in\mathcal{I}. D ∖ A = { α ∈ D : t α ( s α − ζ ) ∈ / V θ + U θ } ∈ I .
So ζ ∈ W I τ − L I M t V θ s \zeta\in\mathrm{WI}_{\tau}-\mathrm{LIM}_{t}^{\mathbf{V}_{\theta}}s ζ ∈ WI τ − LIM t V θ s . Therefore W I τ − L I M t V θ s = R t V θ − I ( Γ s τ ) \mathrm{WI}_{\tau}-\mathrm{LIM}_{t}^{\mathbf{V}_{\theta}}s=\mathcal{R}_{t}^{\mathbf{V}_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau}) WI τ − LIM t V θ s = R t V θ − I ( Γ s τ ) .
Conversely, let W I τ − L I M t V θ s = R t V θ − I ( Γ s τ ) \mathrm{WI}_{\tau}-\mathrm{LIM}_{t}^{\mathbf{V}_{\theta}}s=\mathcal{R}_{t}^{\mathbf{V}_{\theta}}-\mathcal{I}(\Gamma_{s}^{\tau}) WI τ − LIM t V θ s = R t V θ − I ( Γ s τ ) for any net s = { s α } α ∈ D s=\{s_{\alpha}\}_{\alpha\in D} s = { s α } α ∈ D in L L L , with some roughness degree V θ \mathbf{V}_{\theta} V θ . With out any loss of generality we assume that I \mathcal{I} I is not a maximal ideal on D D D . Then there exists a subset A A A of a directed set D D D such that A A A and D ∖ A D\setminus A D ∖ A are both not in I \mathcal{I} I . Let x ≠ θ x\ne\theta x = θ . Then there exists a τ \tau τ -neighbourhood V θ ∗ ∈ N s o l \mathbf{V}_{\theta}^{*}\in\mathcal{N}_{sol} V θ ∗ ∈ N so l such that θ ∈ 2 V θ ∗ \theta\in2\mathbf{V}_{\theta}^{*} θ ∈ 2 V θ ∗ but T x ∉ 2 V θ ∗ Tx\notin2\mathbf{V}_{\theta}^{*} T x ∈ / 2 V θ ∗ , where T = inf α ∈ D ∣ t α ∣ T=\inf_{\alpha\in D}|t_{\alpha}| T = inf α ∈ D ∣ t α ∣ . Let s = { s α } α ∈ D s=\{s_{\alpha}\}_{\alpha\in D} s = { s α } α ∈ D be a net in the following manner:
s α = { θ if α ∈ A , x if α ∈ D ∖ A . s_{\alpha}=\begin{cases} \theta& \text{if } \alpha\in A, \\ x & \text{if } \alpha\in D\setminus A. \end{cases} s α = { θ x if α ∈ A , if α ∈ D ∖ A .
Since V θ ∗ ∈ N s o l \mathbf{V}^{*}_{\theta} \in\mathcal{N}_{sol} V θ ∗ ∈ N so l and T ≤ ∣ t α ∣ T \leq\lvert t_{\alpha}\rvert T ≤ ∣ t α ∣ for all α ∈ D \alpha\in D α ∈ D and T x ∉ 2 V θ ∗ Tx \notin2\mathbf{V}^{*}_{\theta} T x ∈ / 2 V θ ∗ implies t α x ∉ 2 V θ ∗ t_{\alpha}x \notin2\mathbf{V}^{*}_{\theta} t α x ∈ / 2 V θ ∗ . Then
A = { α ∈ D : t α ( s α − θ ) ∈ V θ ∗ + V θ ∗ } = { α ∈ D : t α ( s α − θ ) ∈ V θ ∗ + U θ } , where U θ ⊆ V θ ∗ and U θ ∈ N s o l , \begin{aligned}
A = \{\alpha\in D : t_{\alpha}(s_{\alpha}-\theta) \in\mathbf{V}^{*}_{\theta}+\mathbf{V}^{*}_{\theta}\} \\
&= \{\alpha\in D : t_{\alpha}(s_{\alpha}-\theta) \in\mathbf{V}^{*}_{\theta}+U_{\theta}\}, \text{ where } U_{\theta} \subseteq\mathbf{V}^{*}_{\theta} \text{ and } U_{\theta} \in\mathcal{N}_{sol},
\end{aligned} A = { α ∈ D : t α ( s α − θ ) ∈ V θ ∗ + V θ ∗ } = { α ∈ D : t α ( s α − θ ) ∈ V θ ∗ + U θ } , where U θ ⊆ V θ ∗ and U θ ∈ N so l ,
⟹ D ∖ A = { α ∈ D : t α ( s α − x ) ∈ V θ ∗ + V θ ∗ } = { α ∈ D : t α ( s α − x ) ∈ V θ ∗ + U θ } , where U θ ⊆ V θ ∗ and U θ ∈ N s o l . \begin{aligned}
\Longrightarrow D \setminus A = \{\alpha\in D : t_{\alpha}(s_{\alpha}-x) \in\mathbf{V}^{*}_{\theta}+\mathbf{V}^{*}_{\theta}\} \\
&= \{\alpha\in D : t_{\alpha}(s_{\alpha}-x) \in\mathbf{V}^{*}_{\theta}+U_{\theta}\}, \text{ where } U_{\theta} \subseteq\mathbf{V}^{*}_{\theta} \text{ and } U_{\theta} \in\mathcal{N}_{sol}.
\end{aligned} ⟹ D ∖ A = { α ∈ D : t α ( s α − x ) ∈ V θ ∗ + V θ ∗ } = { α ∈ D : t α ( s α − x ) ∈ V θ ∗ + U θ } , where U θ ⊆ V θ ∗ and U θ ∈ N so l .
Since A A A and D ∖ A D \setminus A D ∖ A not in I \mathcal{I} I . Then x , θ ∈ R t V θ ∗ − I ( Γ s T ) = W I τ − LIM s V θ ∗ x,\theta\in\mathcal{R}^{\mathbf{V}^{*}_{\theta}}_{t}-\mathcal{I}(\Gamma^{\mathcal{T}}_{s})=\mathrm{W}\mathcal{I}_{\tau}-\operatorname{LIM}^{\mathbf{V}^{*}_{\theta}}_{s} x , θ ∈ R t V θ ∗ − I ( Γ s T ) = W I τ − LIM s V θ ∗ . Hence A A A and D ∖ A D \setminus A D ∖ A are both lies in F ( I ) \mathcal{F}(\mathcal{I}) F ( I ) , a contraction since ∅ ∉ F ( I ) \varnothing\notin\mathcal{F}(\mathcal{I}) ∅ ∈ / F ( I ) . So I \mathcal{I} I is an maximal ideal on D D D . □
Corollary 2.22.1. Let I \mathcal{I} I be a non-trival ideal on the directed set D D D , t = { t α } α ∈ D t=\{t_{\alpha}\}_{\alpha\in D} t = { t α } α ∈ D be a weighted net, s = { s α } α ∈ D s=\{s_{\alpha}\}_{\alpha\in D} s = { s α } α ∈ D be a net in a locally solid Riesz space L L L . If ( R t V θ − I ( Γ s T ) ) ∖ ( W I τ − LIM s V θ ) ≠ ∅ \left(\mathcal{R}^{\mathbf{V}_{\theta}}_{t}-\mathcal{I}(\Gamma^{\mathcal{T}}_{s})\right)\setminus\left(\mathrm{W}\mathcal{I}_{\tau}-\operatorname{LIM}^{\mathbf{V}_{\theta}}_{s}\right)\ne\varnothing ( R t V θ − I ( Γ s T ) ) ∖ ( W I τ − LIM s V θ ) = ∅ then I \mathcal{I} I is not a maximal ideal of D D D .
Proof. Let ζ ∈ ( R t V θ − I ( Γ s T ) ) ∖ ( W I τ − LIM s V θ ) \zeta\in\left(\mathcal{R}^{\mathbf{V}_{\theta}}_{t}-\mathcal{I}(\Gamma^{\mathcal{T}}_{s})\right)\setminus\left(\mathrm{W}\mathcal{I}_{\tau}-\operatorname{LIM}^{\mathbf{V}_{\theta}}_{s}\right) ζ ∈ ( R t V θ − I ( Γ s T ) ) ∖ ( W I τ − LIM s V θ ) . Then { α ∈ D : t α ( s α − ζ ) ∈ V θ + U θ } ∉ I \{\alpha\in D:t_{\alpha}(s_{\alpha}-\zeta)\in V_{\theta}+U_{\theta}\}\notin\mathcal{I} { α ∈ D : t α ( s α − ζ ) ∈ V θ + U θ } ∈ / I , for any τ \tau τ -neighbourhood U θ U_{\theta} U θ of θ \theta θ . Since ζ ∉ W I τ − LIM s V θ \zeta\notin\mathrm{W}\mathcal{I}_{\tau}-\operatorname{LIM}^{\mathbf{V}_{\theta}}_{s} ζ ∈ / W I τ − LIM s V θ . Therefore { α ∈ D : t α ( s α − ζ ) ∉ V θ + U θ } ∉ I \{\alpha\in D:t_{\alpha}(s_{\alpha}-\zeta)\notin V_{\theta}+U_{\theta}\}\notin\mathcal{I} { α ∈ D : t α ( s α − ζ ) ∈ / V θ + U θ } ∈ / I . Hence the ideal I \mathcal{I} I is not a maximal ideal on D D D . □
Declarations
Ethical Approval : Not applicable.
Authors’ Contributions : All the authors have equally contributed to this article.
Funding : No funding.
Data Availability : This manuscript has no associated data.
Competing interests : The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.