Description:
Definitions from constructive order theory
Content:
Definitions from constructive order theory} (for forms [0 AT], [0 AU], [1 DH], [1 DI], [1 DJ], [67 AB], [67 AC], [144 B] through [144 M] and Form 413 through Form 416.
Definition: A subset selection \(\cal Z\) is a (class)
function defined on the class of all partial orderings such that if
\((P,\le)\) is a partial ordering then \(\cal Z (P,\le)\) (or \(\cal Z
P\) for short) is a subset of \()\cal P(P)\). The subset selection
\(\cal Z\) is a subset system if for any two partial
orderings \((P_1,\le_1)\) and \((P_2,\le_2)\) if \(f:P_1\to P_2\) is
order preserving then for all \(Z\in \cal Z P_1\), \(f[Z]\in \cal Z
P_2\).
In Ern\'e [2001] the following subset selectors are
considered
- \(\cal A P = \cal P(P)\) (arbitrary subsets of \(P\))
- \(\cal B P = \) one or two element subsets of \(P\)
- \(\cal C P = \) all non-empty chains in \((P,\le)\)
- \(\cal D P = \) all directed subsets in \((P,\le)\)
(\(Z\subseteq P\) is directed by \(\le\) if any two elements of
\(Z\) have a common upper bound.)
- \(\cal E P = \) all one element subsets of \(P\)
- \(\cal F P = \) all finite subsets of \(P\)
- \(\cal W P = \) all non-empty subsets of \(P\) well ordered
by \(\le\).
- \(\cal U P = \) all \(U\subseteq P\) such that \(\forall
x,y\in U\), \(x\lor y\) exists in \(P\) and is in \(U = \) all
\(\bigvee\)-subsemilattices of \(P\).
Note that all of the above except \(\cal U\) are subset systems.
Definition: Assume \((P\le)\) is a partially ordered
set and \(\cal Z\) is a subset selection.
- \((P,\le)\) is \(\cal Z\)-complete (or
\(\cal Z\)-\(\bigvee\)-complete) if every element \(Z\) of \(\cal Z\)
has a \(\sup\) (denoted \(\bigvee Z\) or \(\bigvee_P Z\)).
- A subset \(X\) of \(P\) is \(\cal Z\)-subcomplete if
each \(Z\in \cal Z\cap \cal P(X)\) has a supremum \(s\) in \(P\) that is
contained in \(X\)
- If \(Y\subset P\) then \(s\) is a constructive
supremum of \(Y\) if either \(Y = \emptyset\) and \(s\) is the least
element of \(P\) (if it exists) or \(s\) is an upper bound of \(Y\) and
there is a function \(\psi:P\to Y\) such that \((\forall x\in
P)(s \nleq x \leftrightarrow \psi(x) \nleq x)\).
- \((P,\le)\) is constructively \(\cal Z\)-complete if
each \(Z\in \cal Z P\) has a constructive supremum.
- A constructively complete lattice is a
complete lattice in which every supremum is constructive.
- \(\cal Z^\lor\) is the class function whose domain is the class
of all partial orderings and which is defined by
\[
\cal Z^\lor P = \{ Y\subseteq P : (\forall x\in Y)(\forall z\in
P)(z\le x\to z\in Y)\mathrm{\; and \;}\\ (\forall Z\in \cal P(Y)\cap \cal Z
P)(\mathrm{\; If \;}x =\bigvee_P Y \mathrm{ \; then \;} x\in Y \}
\]
- \((P,\le)\) is constructively directed if there is a
function assigning an upper bound to each non-empty finite subset of \(P\).
- If \(S\) is a set and \(\cal X \subseteq \cal P(S)\) then
\(\cal X\) is \(\cal Z\)-inductive (or \(\cal Z\)-union complete
if for each \(\cal Y\in \cal
Z\cal X\), \(\bigcup \cal Y \in \cal X\).
- \(^c\cal D P\) (or \(^c\cal D (P,\le)\)) is the system of
all constructively directed subsets of \(P\).
- If \(S\) is a set and \(\cal X\subseteq \cal P(S)\) then
\(\cal X\) is a closure system (on \(S\)) if \(\cal X\) is closed
under arbitrary intersections (with \(\bigcap\emptyset = S\)).
- An element \(x\) of a \(\cal Z\)-complete poset \(P\) is
\(\cal Z\)-compact if for all \(Z\in\cal Z P\), if \(x\le \bigvee Z\),
then \(x\in \downarrow Z\). (\(\downarrow Z = \{t\in P : (\exists y\in
Z)(t\le y)\}\). The \()\cal D\)-compact elements are called the compact
elements.
- A \(\cal Z\)-complete poset is \(\cal Z\)-compactly
generated if each of its elements is a supremum of \(\cal
Z\)-compact elements.
- An algebraic lattice is a \(\cal D\)-compactly
generated complete lattice.
- If \(S\) is a set and \(\cal X\subseteq \cal P(S)\) then
\(\cal X\) is a system of finite character for all \(X\),
\(X\in \cal X \Leftrightarrow \cal F X\subseteq \cal X\)
- A Scott closed subset of \(P\) is an element of
\(\cal D^\lor P\).
- A map \(\phi:P\to Q\) between posets preserves
\(\cal Z\)-suprema if for each \(Z\in \cal Z P\) having a supremum
\(s\), the image \(\phi(s)\) is the supremum of \(\phi[Z]\).
- A \(\cal Z\)-frame is a complete lattice in which
the distributive law \(x\land \bigvee Z = \bigvee(x\land Z)\) holds
for all \(x\in P\) and all \(Z\in\cal Z P\).
Howard-Rubin number:
154
Type:
Definitions
Back