Next Article in Journal
A Class of Non-Hopf Bi-Frobenius Algebras Generated by n Elements
Previous Article in Journal
A Chemistry-Based Optimization Algorithm for Quality of Service-Aware Multi-Cloud Service Compositions
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Construction of ε-Nets for the Space of Planar Convex Bodies Endowed with the Banach–Mazur Metric

1
School of Mathematics, North University of China, Taiyuan 030051, China
2
Department of Applied Mathematics, Harbin University of Science and Technology, Harbin 150080, China
*
Author to whom correspondence should be addressed.
These authors contributed equally to this work.
Mathematics 2025, 13(8), 1358; https://doi.org/10.3390/math13081358
Submission received: 12 March 2025 / Revised: 18 April 2025 / Accepted: 19 April 2025 / Published: 21 April 2025
(This article belongs to the Section B: Geometry and Topology)

Abstract

:
In Chuanming Zong’s program to attack Hadwiger’s covering conjecture, which is a long-standing open problem from convex and discrete geometry, the construction of ε -nets for the space of convex bodies endowed with the Banach–Mazur metric plays a crucial role. Recently, Gao et al. provided a possible way of constructing ε -nets for K n , d B M based on finite subsets of Z n theoretically. In this work, we present an algorithm to construct ε -nets for K 2 , d B M and a ( 1 / 4 ) -net for C 2 , d B M is constructed. To the best of our knowledge, this is the first concrete ε -net for C 2 , d B M for such a small ε .

1. Introduction

A compact convex set K R n having interior points is called a c o n v e x   b o d y , whose interior and boundary are denoted by int K and bd K , respectively. Let K n be the set of convex bodies in R n , and let C n be the set of convex bodies in K n that are symmetric about the o r i g i n   o . If A is a set in R n , we denote the c o n v e x   h u l l of A as conv A . Several classical open problems from discrete and convex geometry can be considered via studying certain functionals that are uniformly continuous on ( K n , d B M ) . Take Hadwiger’s covering conjecture for example. Let c ( K ) be the least number of translates of int K needed to cover K.
Conjecture 1
(Hadwiger’s covering conjecture). For each K K n , we have
c ( K ) 2 n ;
the equality holds if and only if K is a parallelotope.
This conjecture has been completely resolved only for the case of n = 2 (cf. [1]) and remains open for n 3 . For further information on this conjecture, we refer to [2,3,4,5,6,7,8,9,10,11,12,13,14].
In fact, c ( K ) equals the minimum number of s m a l l e r   h o m o t h e t i c   c o p i e s (sets having the form x + γ K with x R n and γ ( 0 , 1 ) ) of K needed to cover K. Chuanming Zong introduced a quantitative program to tackle this conjecture via estimating
Γ m ( K ) : = min γ > 0 x i i [ m ] R n s . t . K i [ m ] ( x i + γ K ) ,
where [ m ] = i Z + 1 i m . It is evident that
c ( K ) m Γ m ( K ) < 1 .
The map
Γ m ( · ) : K n [ 0 , 1 ] K Γ m ( K )
is called the covering functional with respect to m. For each m Z + , Γ m ( · ) is an affine invariant. More precisely,
Γ m ( K ) = Γ m ( T ( K ) ) , T A n ,
where A n is the set of non-degenerate affine transformations from R n to R n . It is natural to use the (multiplicative) Banach–Mazur metric d B M M to measure the distance between two convex bodies, where, for each pair of K 1 , K 2 K n , d B M M ( K 1 , K 2 ) is defined as
d B M M ( K 1 , K 2 ) : = min γ 1 x R n , T A n s . t . K 1 T ( K 2 ) γ K 1 + x .
Set
d B M ( K 1 , K 2 ) = ln d B M M ( K 1 , K 2 ) .
It is well known that ( K n , d B M ) is a compact metric space (convex bodies that are affinely equivalent are regarded as the same element, of course).
Let
c n = max Γ 2 n ( K ) K K n .
It is clear that
c ( K ) 2 n , K K n c n < 1 .
Since Γ 2 n ( K ) is continuous on the compact metric space K n , there exists ε > 0 such that, for each K , K K n with d B M ( K , K ) ε , the inequality
| Γ 2 n ( K ) Γ 2 n ( K ) |     1 2 ( 1 c n )
holds.
Let ε be a positive number, and let K 1 , K 2 , , K l ( n , ε ) be l ( n , ε ) convex bodies in K n , where l ( n , ε ) is an integer depending on n and ε . If, for each K K n , there exists a corresponding K i satisfying
d B M M ( K , K i ) 1 + ε ,
then, we call N = { K 1 , K 2 , , K l ( n , ε ) } an ε-net for K n . If, for a given ε > 0 , we can construct an ε-net N and verify that Γ 2 n ( K i ) c n holds for each K i N , then, by (1), we have
Γ 2 n ( K ) 1 2 ( 1 + c n ) < 1 , K K n .
Therefore, Zong proposed the following four steps in [15] to attack Hadwiger’s covering conjecture:
(1)
Obtain a good guess c ^ n of c n by estimating Γ 2 n ( K ) for special classes of convex bodies;
(2)
Choose a suitable ε > 0 ;
(3)
Construct an ε-net N of K n ;
(4)
For each K N , verify that Γ 2 n ( K ) c ^ n .
Clearly, constructing a suitable ε -net N is crucial in this program. Recently, three algorithms have been designed to estimate Γ m ( K ) [2,3,4].
Zong showed in [15] that
l ( n , ε ) 7 n ε c · 14 n · n 2 n + 3 · ε n ,
where c is a universal constant. When n is large and ε is small, the right-hand side is huge. Moreover, one should not expect the cardinality of an ε -net to be small, since there are known lower bounds on the cardinality of such nets, see, e.g., [16,17].
In [18], Shenghua Gao et al. provided a possible way of constructing ε -nets for ( K n , d B M ) based on finite subsets of Z n theoretically. However, there is still no practical algorithm for this purpose. In this paper, we propose an algorithmic method for the first time for constructing ε -nets with a reasonable cardinality in the two-dimensional case. Furthermore, we construct a ( 1 / 4 ) -net for ( C 2 , d B M ) consisting of 30,306 sets. This is the first concrete ε -net for ( C 2 , d B M ) .

2. A General Framework

Definition 1.
Let K Z n and i [ n ] . If
K = conv K Z n ,
then, K is called a convex lattice set. If
x , y K , x y span { e i } [ x , y ] Z n K ,
then K is said to be a lattice set convex in the i-th coordinate.
Clearly, if K is a convex lattice set, then, for each i [ n ] , K is convex at the i-th coordinate.
Proposition 1.
If K Z n is a convex lattice set and H is a closed halfspace of R n , then K H is also a convex lattice set.
Proof. 
We only need to show that conv ( K H ) Z n K H . Actually,
conv ( K H ) Z n ( conv K Z n ) H = K H .
   □
Set
Q n = [ 1 , 1 ] n , n Z + .
For each K K n , let
α ( K ) = max α > 0 T A n s . t . α Q n T ( K ) Q n .
Theorem 1.
Let K K n , M Z + , and α ( 0 , α ( K ) ] . There exists a convex lattice set G M contained in M Q n such that
d B M M ( K , conv ( G M + 1 2 Q n ) ) 1 + 1 M α .
Moreover, if K C n , we may require further G M to be symmetric with respect to o.
Proof. 
The construction is based on the proof of Theorem 6 in [18]. By applying a suitable affine transformation if necessary, we may assume that
α ( K ) Q n K Q n .
Thus α Q n K Q n , which implies that
α M Q n M K M Q n .
Set
G M = z Z n ( z + 1 2 Q n ) M K .
Then
M K G M + 1 2 Q n conv ( G M + 1 2 Q n ) .
For each z G M , there exists a point x M K such that x z 1 2 Q n . Thus, z x + 1 2 Q n . It follows that
G M M K + 1 2 Q n M Q n + 1 2 Q n ,
which implies that
M K conv ( M K + Q n ) = M K + Q n M K + 1 α K = M ( 1 + 1 M α ) K ,
and that (since G M Z n ) G M M Q n . Therefore,
d B M M ( M K , conv ( G M + 1 2 Q n ) ) = d B M M ( K , conv ( G M + 1 2 Q n ) ) 1 + 1 M α .
Next, we show that G M is a convex lattice set. Clearly, K conv K Z n . For each point z conv K Z n , by Carathéodory’s theorem, there exist m [ n + 1 ] , a set of m points { z 1 , , z m } K , and a set of numbers { λ 1 , , λ m } [ 0 , 1 ] such that
z = i [ m ] λ i z i   and   i [ m ] λ i = 1 .
For each i [ m ] , there exists a point x i M K such that z i x i 1 2 Q n . It follows that
( z i [ m ] λ i x i ) = ( i [ m ] λ i ( z i x i ) ) 1 2 Q n .
Since M K is convex, we have
i [ m ] λ i x i M K .
Therefore,
( z + 1 2 Q n ) M K .
Hence, z G M .
In the rest, we consider the case where K C n . Suppose that z G M . Then, there exists a point x 1 2 Q n such that z + x M K , which implies that z x M K = M K . Thus,
( z + 1 2 Q n ) M K ,
which shows that z G M . This completes the proof.    □
Put
α 0 n = inf α ( K ) K K n   and   β 0 n = inf α ( K ) K C n .
Theorem 2
(cf. [19]). Let K K n . Then, the John ellipsoid E (the unique ellipsoid of maximum volume contained in K) of K, whose center is c, satisfies
E K c + n ( E c ) .
Let K K n ; by a suitable affine transformation if necessary, we may assume that the Euclidean unit ball B 2 n is the John ellipsoid of K. From Theorem 2, we have
B 2 n K n B 2 n ,
or, equivalently,
1 n B 2 n 1 n K B 2 n .
Thus,
1 n n Q n 1 n B 2 n 1 n K B 2 n Q n ,
which implies that
α 0 n 1 n n .
Since Q n , C C n , we have
d B M M ( Q n , C ) = min γ 1 T T n s . t . Q n T ( C ) γ Q n ,
where T n is the set of non-degenerate linear transformations from R n to R n . Then
d B M M ( Q n , C ) = min γ 1 T T n s . t . 1 γ Q n T ( C ) Q n .
Therefore,
1 d B M M ( Q n , C ) α ( C ) .
Hence,
inf 1 d B M M ( Q n , C ) C C n inf α ( C ) C C n .
It follows that
β 0 n ( max d B M M ( Q n , C ) C C n ) 1 .
It is proved in [20] that
max d B M M ( Q 2 , C ) C C 2 3 2 .
Thus, β 0 2 2 / 3 .
Corollary 1.
Let ε > 0 , K K n , and C C n .
(1) 
There exists a convex lattice set G 1 contained in
1 α 0 n ε Q n
and containing
α 0 n 1 α 0 n ε Q n Z n
such that
d B M M ( K , G 1 + 1 2 Q n ) 1 + ε .
(2) 
There exists a centrally symmetric convex lattice set G 2 contained in
1 β 0 n ε Q n
and containing
β 0 n 1 β 0 n ε Q n Z n
such that
d B M M ( C , G 2 + 1 2 Q n ) 1 + ε .
Proof. 
We only need to prove (1), and (2) can be proved in a similar way. Clearly, α ( K ) α 0 n . By applying a suitable affine transformation if necessary, we may assume that
α 0 n Q n K Q n .
Let M be the least positive integer such that 1 M α 0 n ε . Then, M = 1 α 0 n ε . As in the proof of Theorem 1, there exists G M M Q n such that
d B M M ( K , conv ( G M + 1 2 Q n ) ) 1 + 1 M α 0 n 1 + ε .
Moreover, α 0 n M Q n M K , which implies that
( α 0 n M Q n ) Z n M K .
For each z Z n , since z M K implies that z G M , we have
( α 0 n M Q n ) Z n = ( α 0 n M Q n ) Z n M K Z n G M .
G 1 : = G M has the desired properties.    □
Theoretically, to construct an ε -net for K n (or C n ), one needs to enumerate all convex lattice sets contained in M 2 Q n and containing M 1 Q n Z n for some positive integers M 1 < M 2 . Then, one needs to take the quotient set of the resulting set with respect to the affine equivalence. Each of these steps is not easy. In the following, we focus on the planar case.

3. Construction of ε -Nets for ( K 2 , d BM )

Let G be the group of transformations whose matrices are
1 0 0 1 , 0 1 1 0 , 1 0 0 1 , 0 1 1 0 , 1 0 0 1 , 1 0 0 1 , 0 1 1 0 , 0 1 1 0 .
Proposition 2.
Let I 1 and I 2 be two subsets of Z 2 . If there exists T G such that I 2 = T ( I 1 ) , then
T ( conv ( I 1 + 1 2 Q 2 ) ) = conv ( I 2 + 1 2 Q 2 ) .
Proof. 
It is clear that, for each T G , T ( 1 2 Q 2 ) = 1 2 Q 2 . From Theorem 3.16 in [5],
conv ( I 2 + 1 2 Q 2 ) = conv ( T ( I 1 ) + T ( 1 2 Q 2 ) ) = conv ( T ( I 1 + 1 2 Q 2 ) ) = T ( conv ( I 1 + 1 2 Q 2 ) ) .
   □
Let M 1 and M 2 be two positive integers satisfying M 1 < M 2 . Let H 0 + and H 0 be the closed halfplanes of R 2 bounded by the line span { e 2 } and containing ( 1 , 0 ) and ( 1 , 0 ) , respectively.
Let K be a convex lattice set contained in M 2 Q 2 and containing M 1 Q 2 Z 2 . Then, both K H 0 + and K H 0 are convex lattice sets. Moreover, K H 0 can be viewed as ( K H 0 + ) , where K is another convex lattice set with the desired properties.
Based on these observations, we can construct ε -nets for ( K 2 , d B M ) with the following steps.
Step 1. Enumerate all subsets G of Z 2 that are convex in the 2nd coordinate and satisfy
M 1 Q 2 Z 2 H 0 + G M 2 Q 2 H 0 + .
Let i [ 0 , M 2 ] Z . Set
C i = ( i , j ) j [ M 2 , M 2 ] Z .
Case 1. i [ M 1 + 1 , M 2 ] . In this case, the intersection of the vertical line passing through ( i , 0 ) and the convex lattice set could be ∅, or a singleton, or a set of the form
[ ( i , β 1 ) , ( i , β 2 ) ] Z 2 ,
with β 1 β 2 . In this case, we set
S i = { } φ ( C i , 1 ) φ ( C i , 2 ) ,
where φ ( A , i ) stands for the collection of all subsets of A consisting of i elements.
Case 2. i 0 , M 1 . In this case, the intersection of the vertical line passing through ( i , 0 ) and the convex lattice set is a set of the form
[ ( i , β 1 ) , ( i , β 2 ) ] Z 2 ,
where β 1 M 1 and β 2 M 1 .
Hence, we set
S i 1 = { } φ ( ( i , j ) M 1 < j M 2 , j Z , 1 )
and
S i 2 = { } φ ( ( i , j ) M 2 j < M 1 , j Z , 1 ) .
We do not consider the situation j = ± M 1 , since we take the union of the resulting set with M 1 Q 2 Z 2 later.
The set is
P = ( i = 0 M 1 ( S i 1 × S i 2 ) ) × i = M 1 + 1 M 2 S i .
For each member
P = ( S 0 1 , S 0 2 , , S M 1 1 , S M 1 2 , S M 1 + 1 , , S M 2 ) P ,
we can recover a lattice set that is convex in the 2nd coordinate in the following way. A denotes the cardinality of a set A.
Let i [ 0 , M 2 ] Z .
Case 1. i [ M 1 + 1 , M 2 ] .
  • If S i 1 , then we set T i = S i .
  • If S i has the form { ( i , β 1 ) , ( i , β 2 ) } , then we set T i = [ ( i , β 1 ) , ( i , β 2 ) ] Z 2 .
Case 2. i 0 , M 1 . If S i 1 = , we set β 1 = M 1 ; otherwise, let β 1 be the 2nd coordinate of the point in S i 1 . Similarly, if S i 2 = , we set β 2 = M 1 ; otherwise, we let β 2 be the 2nd coordinate of the point in S i 2 . Let T i = [ ( i , β 1 ) , ( i , β 2 ) ] Z 2 .
The set T ( P ) = i [ 0 , M 2 ] T i is a lattice set convex in the 2nd coordinate, see Algorithm 1.
Algorithm 1 Convert P to set T ( P ) .
Require
P P
Ensure
T ( P ) , a lattice set convex in the 2nd coordinate generated by P.
   1:
Initialize a set T ( P )
   2:
for   i [ 0 , M 1 ]   do
   3:
     S i 1 P [ 2 i ] P [ i ] denotes the i-th elements of P
   4:
     S i 2 P [ 2 i + 1 ]
   5:
    if  S i 1 =  then
   6:
         u b M 1
   7:
    else
   8:
         u b β 1 ▹ Second coordinate of the point in S i 1
   9:
    end if
 10:
    if  S i 2 =  then
 11:
         l b M 1
 12:
    else
 13:
         l b β 2 ▹ Second coordinate of the point in S i 2
 14:
    end if
 15:
     T i { i } × [ l b , u b ]
 16:
     T ( P ) T ( P ) T i
 17:
end for
 18:
for   i [ M 1 + 1 , M 2 ]  do
 19:
     S i P [ i + M 1 + 1 ]
 20:
    if  S i 1  then
 21:
         T ( P ) T ( P ) S i
 22:
    else if  S i = 2  then
 23:
         l b min x ( i , x ) P [ i + M 1 + 1 ]
 24:
         u b max x ( i , x ) P [ i + M 1 + 1 ]
 25:
         T i { i } × [ l b , u b ]
 26:
         T ( P ) T ( P ) T i
 27:
    end if
 28:
end for
 29:
return   T ( P )
Step 2. Obtain all convex lattice sets satisfying
M 1 Q 2 Z 2 G M 2 Q 2 .
The case for K 2 : Let L 0 and L be an empty list. For each P P , if T ( P ) is a convex lattice set, then we add T ( P ) to L 0 ; otherwise, we ignore P. Let ( T 1 , T 2 ) L 0 × L 0 . If T 1 T 2 is a convex lattice set, then we add T 1 T 2 to L .
The case for C 2 : Let L be a new empty list. For each P P , if T ( P ) ( T ( P ) ) is a convex lattice set, we add T ( P ) ( T ( P ) ) to L ; otherwise, we ignore P.
For each P P , Algorithm 2 is used to check whether T ( P ) is a lattice set convex in the 1st coordinate. Thereafter, Algorithm 3 is used to obtain all convex lattice sets in T ( P ) = T ( P ) P P .
Algorithm 2 Checking convexity in the 1st coordinate for a lattice set.
Require:
T ( P ) T ( P )
   1: 
for   i [ M 2 , M 2 ]   do
   2:
     Y i x ( x , i ) T ( P )
   3:
    if  Y i 2  then
   4:
         l b min ( Y i )
   5:
         u b max ( Y i )
   6:
        if  Y i < u b l b + 1  then
   7:
           return False
   8:
        end if
   9:
    end if
 10:
end for
 11:
return True
Algorithm 3 Obtain all convex lattice sets in T ( P ) .
Require:
P
   1:
L 0
   2:
for   P P   do
   3:
     S 0 to _ set ( P ) ▹ Algorithm 1
   4:
    if is_x_convex ( S 0 )   then ▹ Algorithm 2
   5:
         S conv ( S 0 )
   6:
        if  S 0 = ( S Z 2 )  then
   7:
            L 0 L 0 S
   8:
           continue
   9:
        end if
 10:
    end if
 11:
end for
 12:
return   L 0
Step 3. Let L be the list obtained in the last step. For each set A L , remove every element B L { A } for which there exist T G and x R 2 such that
T ( A ) + x = B ,
to obtain a new list L , see Algorithm 4.
Algorithm 4 Removing redundant members in L .
Require:
L , G
   1:
L
   2:
S
   3:
for   A L  do
   4:
    Compute the centroid c A of conv A
   5:
     A = conv A c A
   6:
    if  A S  then
   7:
        continue
   8:
    end if
   9:
     L L { A }
 10:
    Set G ( A ) = T ( A ) T G
 11:
     S S G ( A )
 12:
end for
 13:
return   L
Step 4. Constructing the ε -net.
  • For each set A L :
    Generating the convex body conv ( A + ( 1 / 2 ) Q 2 ) . Clearly, we only need to store the set of vertices of each convex polytope.
    Temporarily store the results in a new list L temp .
  • Update the list L by replacing it with L temp .
The updated list L is a desired ε -net.
From [15] (Theorem A), if d B M M K , L 1 + ε holds for some positive number ε , then
| Γ m ( K ) Γ m ( L ) | ε .
This implies that, if one can construct an ε -net N for ( K n , d B M ) with
ε < 1 sup Γ 2 n ( L ) L N ,
then,
Γ 2 n ( K ) sup Γ 2 n ( L ) L N + ε < 1 , K K n .
In [21], M. Lassak proved that
Γ 4 ( K ) 2 2
holds for each planar convex body K, and this bound can be attained.
Therefore, to confirm Hadwiger’s covering conjecture for C 2 , we need to construct an ε-net of ( C 2 , d B M ) for some ε ( 0 , 1 2 / 2 ) .
If we take M 2 = 6 and M 1 = 4 , then, from Corollary 1, the set of all convex lattice sets contained in M 2 Q 2 and containing M 1 Q 2 Z 2 is a ( 1 / 4 ) -net for ( C 2 , d B M ) . In the following, we construct such a net.
For i { 5 , 6 } , we have
S i = { } φ ( ( x , y ) x { 5 , 6 } , y [ 6 , 6 ] Z , 1 ) φ ( ( x , y ) x { 5 , 6 } , y [ 6 , 6 ] Z , 2 ) .
For i [ 0 , 4 ] Z , we have
S i 1 = { } φ ( { ( i , 5 ) , ( i , 6 ) } , 1 )
and
S i 2 = { } φ ( { ( i , 5 ) , ( i , 6 ) } , 1 ) .
Since
S i 1 = S i 2 = 3
and S i = 1 + 13 + C 13 2 = 92 , we have
P = ( 3 × 3 ) 5 × ( 1 + 13 + C 13 2 ) 2 = 499 , 790 , 736 .
Thus, T ( P ) = 499 , 790 , 736 . By Step 2, there are 434,849 members in T ( P ) that are convex lattice sets. After applying Algorithm 4, there are 30,306 sets left. Finally, we obtain the convex hull of the Minkowski sum of these convex lattice sets and ( 1 / 2 ) Q 2 as in Step 4. The family of these convex lattice sets is a ( 1 / 4 ) -net for ( C 2 , d B M ) .
In [4], Han et al. proposed an algorithm based on CUDA for estimating covering functionals of two-dimensional convex polytopes. On a computer equipped with an AMD Ryzen 9 3900X 12-core processor, and the NVIDIA A4000 graphics processor, and taking B 2 2 for example, this algorithm takes approximately 0.217 s to obtain an estimation of Γ 4 ( B 2 2 ) that is no greater than 0.75 . Thus, we can complete the verification of the 30,306 two-dimensional convex bodies in this ( 1 / 4 ) -net for ( C 2 , d B M ) within approximately 2 h and confirm that Γ 4 ( K ) Γ 4 ( conv ( G M + ( 1 / 2 ) Q 2 ) ) + 1 / 4 < 1 .
Remark 1.
An ε-net N for ( K n , d B M ) or ( C n , d B M ) is said to be reduced if no two members in N are affinely equivalent. Most likely, the ( 1 / 4 ) -net we obtained is not reduced. Clearly, for each member L of this ( 1 / 4 ) -net, 2 L is the convex hull of a convex lattice set. There are several recent results for deciding whether two convex lattice sets are affinely equivalent; see, e.g., [22]. To the best of our knowledge, there is no very efficient algorithm for this purpose yet.

Author Contributions

Conceptualization, S.W.; methodology, Y.C., S.W. and Y.L.; software, S.W. and Y.L.; validation, S.W.; formal analysis, Y.C., S.G. and S.W.; data curation, Y.L.; writing—original draft preparation, Y.C., S.G. and Y.L.; writing—review and editing, S.W.; funding acquisition, S.W. All authors have read and agreed to the published version of this manuscript.

Funding

The authors were supported by the National Natural Science Foundation of China (grant numbers 12071444 and 12201581) and the Fundamental Research Program of Shanxi Province (grant numbers 202403021221109, 20210302124657, 202103021224291, and 202303021221116).

Data Availability Statement

Implementations of the main algorithms discussed in this paper, along with the results of the generated 30,306 convex lattice sets, are available at [23], accessed on 24 February 2025.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Levi, F.W. Überdeckung eines Eibereiches durch Parallelverschiebung seines offenen Kerns. Arch. Math. 1955, 6, 369–370. [Google Scholar] [CrossRef]
  2. He, C.; Lv, Y.; Martini, H.; Wu, S. A branch-and-bound approach for estimating covering functionals of convex bodies. J. Optim. Theory Appl. 2023, 196, 1036–1055. [Google Scholar] [CrossRef]
  3. Yu, M.; Lv, Y.; Zhao, Y.; He, C.; Wu, S. Estimations of Covering Functionals of Convex Bodies Based on Relaxation Algorithm. Mathematics 2023, 11, 2000. [Google Scholar] [CrossRef]
  4. Han, X.; Wu, S.; Zhang, L. An Algorithm Based on Compute Unified Device Architecture for Estimating Covering Functionals of Convex Bodies. Axioms 2024, 13, 132. [Google Scholar] [CrossRef]
  5. Soltan, V. Lectures on Convex Sets; World Scientific Publishing Co. Pte. Ltd.: Hackensack, NJ, USA, 2015. [Google Scholar] [CrossRef]
  6. Bezdek, K.; Khan, M.A. The geometry of homothetic covering and illumination. In Discrete Geometry and Symmetry; Springer: Cham, Switzerland, 2018; Volume 234. [Google Scholar] [CrossRef]
  7. Boltyanski, V.; Martini, H.; Soltan, P.S. Excursions into Combinatorial Geometry, Universitext; Springer: Berlin/Heidelberg, Germany, 1997. [Google Scholar] [CrossRef]
  8. Lassak, M. Solution of Hadwiger’s covering problem for centrally symmetric convex bodies in E3. J. Lond. Math. Soc. 1984, 2, 501–511. [Google Scholar] [CrossRef]
  9. Martini, H.; Soltan, V. Combinatorial problems on the illumination of convex bodies. Aequationes Math. 1999, 57, 121–152. [Google Scholar] [CrossRef]
  10. Papadoperakis, I. An estimate for the problem of illumination of the boundary of a convex body in E3. Geom. Dedicata 1999, 75, 275–285. [Google Scholar] [CrossRef]
  11. Rogers, C.A.; Zong, C. Covering convex bodies by translates of convex bodies. Mathematika 1997, 44, 215–218. [Google Scholar] [CrossRef]
  12. Yu, M.; Gao, S.; He, C.; Wu, S. Estimations of covering functionals of simplices. Math. Inequal. Appl. 2023, 26, 793–809. [Google Scholar] [CrossRef]
  13. He, C.; Martini, H.; Wu, S. On covering functionals of convex bodies. Math. Anal. Appl. 2016, 2, 1236–1256. [Google Scholar] [CrossRef]
  14. Li, X.; Meng, L.; Wu, S. Covering functionals of convex polytopes with few vertices. Arch. Math. 2023, 119, 135–146. [Google Scholar] [CrossRef]
  15. Zong, C. A quantitative program for Hadwiger’s covering conjecture. Sci. China Math. 2010, 53, 2551–2560. [Google Scholar] [CrossRef]
  16. Rudelson, M. On the complexity of the set of unconditional convex bodies. Discret. Comput. Geom. 2016, 55, 185–202. [Google Scholar] [CrossRef]
  17. Pisier, G. On the metric entropy of the Banach-Mazur compactum. Mathematika 2015, 61, 179–198. [Google Scholar] [CrossRef]
  18. Gao, S.; Martini, H.; Wu, S.; Zhang, L. New covering and illumination results for a class of polytopes. Arch. Math. 2024, 122, 599–607. [Google Scholar] [CrossRef]
  19. John, F. Extremum problems with inequalities as subsidiary conditions. In Studies and Essays Presented to R; Interscience Publishers: New York, NY, USA, 1948. [Google Scholar] [CrossRef]
  20. Stromquist, W. The maximum distance between two-dimensional Banach spaces. Math. Scand. 1981, 48, 205–225. [Google Scholar] [CrossRef]
  21. Lassak, M. Covering a plane convex body by four homothetical copies with the smallest positive ratio. Geom. Dedicata 1986, 21, 157–167. [Google Scholar] [CrossRef]
  22. Cai, Z.; Zhang, Y.; Liu, Q. On the classification of lattice polytopes via affine equivalence. arXiv 2024, arXiv:2409.09985. [Google Scholar]
  23. Lyu, Y. Computer-Assisted Proofs [Source Code]. Available online: https://github.com/lyuyunfang/Construction-of-varepsilon–nets-for-the-space-of-planar-convex-bodies (accessed on 24 February 2025).
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Chen, Y.; Lyu, Y.; Gao, S.; Wu, S. Construction of ε-Nets for the Space of Planar Convex Bodies Endowed with the Banach–Mazur Metric. Mathematics 2025, 13, 1358. https://doi.org/10.3390/math13081358

AMA Style

Chen Y, Lyu Y, Gao S, Wu S. Construction of ε-Nets for the Space of Planar Convex Bodies Endowed with the Banach–Mazur Metric. Mathematics. 2025; 13(8):1358. https://doi.org/10.3390/math13081358

Chicago/Turabian Style

Chen, Yanmei, Yunfang Lyu, Shenghua Gao, and Senlin Wu. 2025. "Construction of ε-Nets for the Space of Planar Convex Bodies Endowed with the Banach–Mazur Metric" Mathematics 13, no. 8: 1358. https://doi.org/10.3390/math13081358

APA Style

Chen, Y., Lyu, Y., Gao, S., & Wu, S. (2025). Construction of ε-Nets for the Space of Planar Convex Bodies Endowed with the Banach–Mazur Metric. Mathematics, 13(8), 1358. https://doi.org/10.3390/math13081358

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop