On the existence of \(3\)-way \(k\)-homogeneous Latin trades (Q1759386)

From MaRDI portal





scientific article
Language Label Description Also known as
default for all languages
No label defined
    English
    On the existence of \(3\)-way \(k\)-homogeneous Latin trades
    scientific article

      Statements

      On the existence of \(3\)-way \(k\)-homogeneous Latin trades (English)
      0 references
      0 references
      0 references
      0 references
      20 November 2012
      0 references
      This paper deals with the problem of the existence and construction of 3-way \(k\)-homogeneous Latin trades. Every Latin square \(L\) of order \(n\) can be represented as a subset of \(S \times S \times S\) where \(S= \{1,2,\dots,n\}\), i.e., \(L=\{(i, j, k)\mid\) element \(k\) occurs in position \((i, j)\}\). It is called a partial Latin square \(P\) of order \(n\) if each element of \(S\) appears at the most once in each row and in each column. The set \(C_p = \{(i, j) \mid (i, j, k) \in P\}\) is called the shape of \(P\), and \(|C_P|\) is called the volume of \(P\). A is \(\mu\)-way Latin trade of volume \(s\) is made up of \(\mu\) partial Latin squares, denoted by \((T_1,T_2,\dots, T_\mu)\) containing exactly the same \(s\) filled cells and satisfying the following condition: {\parindent=7mm \begin{itemize}\item[--]if cell \((i, j)\) is filled it contains a different entry in each of the \(\mu\) partial Latin squares, and such that row \(i\) in each of the \(\mu\) partial Latin squares contains the same symbols, set-wise, and column likewise. \end{itemize}} A \(\mu\)-way Latin trade is called a \(\mu\)-way \(K\)-homogenous Latin trade, denoted by \((\mu,k,m)\) Latin trade with \(\mu\leq k\), if it is obtained from another one by deleting its empty rows and empty columns, and if in each row and each column, \(T_r\) \((1\leq r\leq\mu)\) contains exactly \(k\) elements, and each element appears in \(T_r\) exactly \(k\) times. In this paper, some general constructions for \((\mu,k,m)\) Latin trades have been presented with special emphasis on the case when \(\mu=3\). The following is the main theorem in this paper: Theorem: All \((3,k,m)\) Latin trades (\(m \geq k \geq 3\)) exist for {\parindent=7mm \begin{itemize}\item[(a)]\(k=4\), except for \(m=6\) and \(7\), and possibly for \(m=11\); \item[(b)]\(k=5\) except possibly for \(m=6\); \item[(c)]\(6\leq k\leq 13\); \item[(d)]\(k=15\); \item[(e)]\(k\geq 4\) and \(m\geq k^2\); \item[(f)]\(m=5l\) (where \(l\) is a positive integer, except \(l=6\)); \item[(g)]\(m=7l\) (\(l\) being a positive integer, except \(l=6\)), for \(k=4\) and \(l=1\). \end{itemize}}
      0 references
      0 references
      Latin square
      0 references
      Latin trade
      0 references
      \(\mu\)-way Latin trade
      0 references
      \(\mu\)-way \(k\)-homogeneous Latin trade
      0 references

      Identifiers