Unavoidable pattern

This is an old revision of this page, as edited by Xiaoshiqi2008 (talk | contribs) at 14:33, 18 February 2020 (adding the graph pattern section, fixed some typos). The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.

In mathematics and theoretical computer science, a pattern(term) is a sequence of characters over some alphabet .

A word  is an instance of pattern , if there exists a non-erasing semigroup morphism  such that . Non-erasing means .

A word matches(encounters) pattern if there exists a factor of , which is also an instance of . Otherwise, avoids . If avoids , is also called -free.

A pattern is unavoidable on a finite alphabet if there exists , word matches for all . Otherwise, is avoidable on , which implies there exist infinitely many words over alphabet that avoid . This is equivalent to saying that is avoidable on if there exists an infinite word that avoids .

A pattern is an unavoidable pattern(blocking term) if is unavoidable on any finite alphabet.

When a pattern is said to be unavoidable and does not specify an alphabet, then is unavoidable for any finite alphabet. Conversely, a pattern is said to be avoidable if is avoidable on some finite alphabet.

Zimin words

Given alphabet  , Zimin words(patterns) are defined recursively   for   and base case  .

A word   is unavoidable if and only if it's a factor of a Zimin word.[1]


Given a finite alphabet  , let   representing the smallest   such that   matches   for all  . We have following properties:[2]

  •  
  •  
  •  
  •  

Given a pattern  , let   representing the number of distinct characters of  .   is avoidable if  . Hence   is one of the longest unavoidable patterns constructed by alphabet  .

Pattern reduction

Given a pattern   over some alphabet  , we say   is free for   if there exist subsets   of   such that the following hold:

  1.   is a factor of   and   ↔    is a factor of   and  
  2.  

e.g. let  , then   is free for   since there exist   satisfied conditions above.

A pattern   can reduce to pattern   if there exists a character   such that   is free for  , and   can be obtained by removing all occurrence of   from  . Denote as  .

e.g. let  , then   can reduce to   since   is free for  .

Given patterns  , if   can reduce to   and   can reduce to  , then   can reduce to  . Denote as  .

A pattern   is unavoidable if and only if   can reduce to the empty word, hence  .[3][1]

Avoidability index

A pattern   is  -avoidable if   is avoidable on an alphabet   of size  . Otherwise,   is  -unavoidable, which means   is unavoidable on every alphabet of size  .[4][5]

if pattern   is  -avoidable, then   is  -avoidable for all  .

The avoidability index of a pattern   is the smallest   such that   is  -avoidable,   if   is unavoidable.[6]

  • Many patterns are 2-avoidable.
  • xx,xxy,xyy,xxyx,xxyy,xyxx,xyxy,xyyx,xxyxx,xxyxy,xyxyy, as well as many doubled words, are 3-avoidable, though this list is not complete.[7]
  • abwbaxbcyaczca is 4-avoidable, as well as other locked words. (Baker, McNulty, Taylor 1989)
  • abvbawbcxacycdazdcd is 5-avoidable. (Clark 2004)
  • no words with index greater than 5 have been found.

Graph pattern avoidance

Given a graph  , a coloring   for the edges   is said to match the pattern   if there exists a simple path   in   such that the sequence of colors   matches  . Otherwise,   is said to avoid   or  -free.[8]

The pattern chromatic number   is the minimal number of distinct colors needed for a  -free coloring   over the graph  .

Examples

  • The Thue–Morse sequence is cube-free and overlap-free, hence it avoids the patterns   and  .[4][5]
  • The patterns   and   are unavoidable on any alphabet, since they are factors of the Zimin words.[9][10]
  • The power patterns   for   are 2-avoidable.[4]
  • A square-free word is one avoiding the pattern  . An example is the word over the alphabet   obtained by taking the first difference of the Thue–Morse sequence.[11][12] See also Square-free word.
  • all binary patterns can divided into three categories:[7]
    •   are unavoidable.
    •   are 3-avoidable
    • others are 2-avoidable

References

  1. ^ a b Zimin, A. I. (1984). "BLOCKING SETS OF TERMS". Mathematics of the USSR-Sbornik. 47 (2): 353. doi:10.1070/SM1984v047n02ABEH002647. ISSN 0025-5734.
  2. ^ Joshua, Cooper; Rorabaugh, Danny (2013). Bounds on Zimin Word Avoidance. arXiv:1409.3080. Bibcode:2014arXiv1409.3080C.
  3. ^ BEAN, DWIGHT RICHARD; EHRENFEUCHT, ANDRZEJ; MCNULTY, GEORGE FRANK (1979). Avoidable Patterns in Strings of Symbols (PDF).
  4. ^ a b c Lothaire (2011) p. 113
  5. ^ a b Berstel et al (2009) p.127
  6. ^ Lothaire (2011) p.124
  7. ^ a b Lothaire, M. (2002). Algebraic Combinatorics on Words.
  8. ^ Grytczuk, Jarosław (2007-05-28). "Pattern avoidance on graphs". Discrete Mathematics. The Fourth Caracow Conference on Graph Theory. 307 (11): 1341–1346. doi:10.1016/j.disc.2005.11.071. ISSN 0012-365X.
  9. ^ Allouche & Shallit (2003) p.24
  10. ^ Lothaire (2011) p.115
  11. ^ Pytheas Fogg (2002) p.104
  12. ^ Berstel et al (2009) p.97