different definitions of a subnet

162 Views Asked by At

The classical definition of subnet seems to be that $\Psi: J\to X$ is a subnet of $\Phi: I\to X$ if there exists a monotone, final map $h: J\to I$ s.t. $\Psi = \Phi\circ h$. I found another definition that states that $\Psi$ is a subnet of $\Phi$ if for every subset $A\subset X$ holds that $\Psi$ is eventually in $A$ whenever $\Phi$ is eventually in $A$. Here $\Phi$ eventually in $A$ means that there exists $i\in I$ s.t. $\Phi(k)\in A$ for $k\geq i$.

The second definition has the advantage that it is better suited to translate between nets and filters.

My question is: beside the point with the filters, are there differences between those two definitions with "real" impact?

1

There are 1 best solutions below

0
On BEST ANSWER

Probably not. This PDF of notes by Martin Sleziak compares the three common definitions of subnet; you’ve given the definitions of what are there called AA-subnets, the most general type, and Willard subnets, the least general. The final result is that if $\mu=\langle y_\beta:\beta\in\Bbb B\rangle$ is an AA-subnet of a net $\nu=\langle x_\alpha:\alpha\in\Bbb A\rangle$ in a set $X$, then $\mu$ is AA-equivalent to a Willard subnet of $\nu$, meaning that $\nu$ has a Willard subnet that generates the same eventuality filter on $X$ as $\mu$. (The eventuality filter of $\mu$ is just the family of subsets of $X$ that contain a tail set of $\mu$.)

Thus, the choice of type of subnet is should be just a matter of convenience for just about any result that can be stated in terms of filters.