>>6916

Nice work!

Here's my working idea for order of operations.

Start at c (of course!)

Create (1,c)

Create na transform records for (e,1) and (-f,1)

Then begin expanding outward from the na transform records.

We already have a starting t value from the na transforms

We know an and a(n-1) have a lower t value

We know bn and b(n-1) have a larger t value

We also know that at the correct t values, we can move from an to bn, and they are [t+n] elements apart

We also know that at the correct t value, a(n-1) and b(n-1) are [t+n-1] elements apart

So we “iterate” (not really, because we’re moving by multiples and factors), looking for the 4 elements that match and give us our an, bn, a(n-1), and b(n-1) values.

It’s definitely an o(log t) way to search

Thoughts, Anons?

>>6920

Ima smoke a toke to that! Dank. Nice work GA!

>>6919

By creating only the needed records around each na transform record. Surrounding (-f,1) and (e,1)

>>6919

and looking for the an, ab, a(n-1), and b(n-1) records. They occurs in a semi predictable pattern surrounding (na transform).

We're searching the factor tree now, niggaz.

All that "iterate" stuff pales in comparison.