hello /sci/ the longest word composed by 'a' and 'b'

>>7838787
your example has the string "aabbaabbaab" repeat continuously.

what is the minimum number of characters required to be a pattern?

a more technical formulation of my question would be "is the quotient of the free group [math]F_2[/math] by [math]\forall x~x^3=e[/math] finite?"

>>7838797
look again at my example
also, the "..." do not imoly that the pattern is repeating, it means that i could go much further but it gets boring

any repeated sequence of characters is a pattern

>>7838798
>free group
Don't you mean the free monoid? The free group would have 4 generators, including a^-1 and b^-1.

>>7838805
if i state that [math]\forall x~x^3=e[/math] it means that there is no need for [math]a^{-1}[/math] since it is the same as [math]aa[/math]

>>7838811
Yeah, but I meant the object you have before factoring out this relation. You're taking the quotient of the free monoid modulo x^3=e.

>>7838814
it is the exact same thing after the quotient anyway
make it a monoid if you prefer

op here

i am starting to become pretty sure it is infinite
consider the following sequence
[math]x_0=ab[/math]
[math]x_{n+1}=x_n x_n {x_n}^{c} {x_n}^{c}[/math]
where the ^c thing means "swap a and b"

it yields an infinite "word"
at least im pretty sure it does, trying to write a proof

>>7838861

Why do you need to repeat x_n ? It seems like a similar argument could be made for the sequence

x_0 = a
x_{n+1} = (x_n) (x_n)^c

>>7838814
>>7838805
Free group on...
implies an identity and inverse. You can think of it as taking those generators and applying the group axioms to them. You could add generators a^-1 and b^-1, but then it wouldn't be free because you've imposed this extra relation - that a^-1 is the inverse of a.

>>7838931
I worded that a bit poorly. The group would be still be free but it wouldn't be the free group on the set {a,b,a^-1,b^-1}. The free group on that set of generators is isomorphic to the free group on {a,b,c,d},.

>>7838913
please stop writing weeaboo text faces this is a science board

>>7838787
abaa repeats a pattern?

And aabbaabb doesn't repeat "aabb"?

I think you need to define your terms better.

>>7838861
If i got your definition right i get:
ab
ababbaba
ababbabaabab > babababa < abab

what do you mean by swap a and b.
also, this is such a good question.
>>7839119
abaa repeat a.
after that we consider 3 repetition's.

>>7839155
aba repeats a as well. Again, what do you mean by repeating a pattern?

>>7839155
damn youre right
thanks for pointing that out
>>7839172
three consecutive occurence of a "word"

>>7839155
>>7838913
also it seems it doesnt happen with this one
this is no proof though

>>7839214
still op here
i have a sketch of a proof that might work with
[math]x_0=a[/math]
[math]x_{n+1}=x_n {x_n}^c[/math]
which yields abbabaabbaababbabaababbaabbabaab...
(but keep in mind that i was pretty sure my previous solution worked too so there might be some oversights)

lets use a technical lemma:
you can prove by induction that if you only consider the letters in position [math]p_0 + 2^k n[/math] then the pattern is the same as the starting pattern (maybe shifted or "swapped" but that doesnt matter)

then here is the draft of my proof:
first, notice that the letters are grouped by pairs, either 'ab' or 'ba'
therefore, there cant be two consecutive occurences of a pattern of odd length: if the first occurence has k 'a' and k+1 'b', then the next one will have k+1 'a', and k 'b'
(that forbids patterns of length 1, 3, 5,
...)

then we notice that that proof still holds for patterns of length [math]2^k n[/math] (with n odd) if we use our lemma and only consider every 2^k th letter starting with the first one of the supposed pattern

therefore there is no repeating pattern, regardless of length

https://en.wikipedia.org/wiki/Thue%E2%80%93Morse_sequence