428 Combinatorics of Compositions and Words
for(n=1;n<11;n++){
numberword=0; num2=0;
for(i=0;i<n;i++) word[i]=0;
builtwords(0,lenp);
printf("The number of words in [%d]^%d that avoid the
pattern is 10^7*%d+%d\n",NK,n,num2,numberword);
fprintf(fout,"The number of words in [%d]^%d that avoid the
pattern is 10^7*%d+%d\n",NK,n,num2,numberword);
}
printf("\n");
fprintf(fout,"\n");
}
}
else
for(j=1;j<=lenp;j++){
p[l]=j; builtpatterns(l+1,lenp);
}
}
void main()
{
int lenp=3; /* length of the subsequence patterns */
fout=fopen("data.dat","w");
builtpatterns(0,lenp);
fclose(fout);
}
We have listed the nontrivial Wilf-equivalence classes for subsequence pat-
terns of length four and five for words in Tables 6.3 and 6.4. Propositions 6.18
and 6.19 together with Theorem 6.20 completely classify the Wilf-equivalence
classes for these patterns. Tables G.3 and G.4 contain all Wilf-equivalence
classes, including the ones that consist of a single pattern. Due to the much
larger number of words of length n (as opposed to compositions of n), we
present the values AW
τ
[5]
(n)forn =4,...,9andAW
τ
[6]
(n)forn =5, 6,...,9,
respectively. If these values do not distinguish between two patterns, then we
provide values obtained using the above program for larger n from a conve-
nient alphabet to show that the patterns are indeed in different classes.
TABLE G.3: Subsequence patterns of length four
τ The sequence {|AW
τ
[5]
(n)|}
9
n=4
|AW
τ
[6]
(9)|
1111 620, 3020, 14300, 65100, 281400, 1138200
1112,1121 615, 2935, 13425, 58259, 237798, 906722
1122 615, 2915, 13095, 55235, 217710, 799910
1123,1132 615, 2935, 13475, 59319, 250306, 1013790
© 2010 by Taylor and Francis Group, LLC