2 ms·
> ... the article doesn't clearly describe > how to construct said sequence ... I thought the construction was clear. In particular, this argument does not
by ColinWright 13y ago
> ... the article doesn't clearly describe
> how to construct said sequence ...
I thought the construction was clear. In particular, this argument does not use Cantor's construction wherein the interval end-points are always algebraics. Instead, it ...
* Finds the next listed algebraic a in the current interval,
* divides the current interval into three, and
* chooses the left-most sub-interval that neither contains nor touches a.
The sequence is then the left-most endpoints of the intervals. In particular, the sequence is composed entirely of rationals. More, the argument works even if the set being avoided (which must be countable, but need not be the algebraics) is not dense.