I No structure in ##(d,x)## in ##\textbf{Met*}(X)## is admissible

elias001
Messages
359
Reaction score
24
TL;DR Summary
I want to know how to show no structure ##(d,x)## in ##\textbf{Met*}(X)## such that each ##\pi_i:(X,d,x)\to (\textbf{R}_+, |x-y|,0)## is admissible
The following question are taken from ##\textit{Arrows, Structures and Functors the categorical imperative}## by Arbib and Manes, and from ##\textit{Algebraic Theories}## by Manes

##\color{blue}{Question}\color{blue}{/difficulties:}##

I am having a lot of difficulties doing the ##\color{green}{Exercise:}## from the end of the second quoted passage below for a few reasons. The ##\textbf{difficulties}## I have are as follows:

I don't know how to describe the mapping ##\pi_i:(X,d,x)\to (\textbf{R}_+, |x-y|,0)## written out in explicit functional elements notation. From ##X=\textbf{R}^I## with ##I## being infinite, the domain of ##\pi_i## would involved some sort of product of metrics, and ##\pi_i## is suppose to serve as some sort of projection map. So we would have ##d_1(a_1,b_1), d_2(a_2,b_2), \ldots, d_i(a_i,b_i), i\in I,## and ##\pi_jd_i(a,b)=d_j(a_j,b_j),## where each ##d_i(a_i,b_i),## is a metric in ##\textbf{Met*}(X)## for all ##i\in I,## ##d_j(a_j,b_j)## is a metric in ##\textbf{Met*}(\textbf{R}_+)## and both ##a=(a_i)_{i\in I},b=(b_i)_{i\in I},## are elements of ##X=\textbf{R}.## I am not sure if this interpretation is correct.

In that exercise question, the author asks the reader to show that there is no structure in ##(d,x)## that would make each ##\pi_i## admissible. I think I am suppose to do it by contradiction and assume such added structure exists, then either show that ##\pi_i## is either not a contraction or is not an isometry. I am not sure if this would be a feasible approach.

##\color{purple}{Background}## ##\color{purple}{Information}##

(From Manes)

##\textbf{Categories of}## ##\mathscr{K-}####\textbf{Objects with Structure.}## Let ##\mathscr{K}## be a (fixed base) category. A ##\textit{literal category,}## ##\mathscr{C,}## of ##\mathscr{K-}####\textit{objects with structure}## is defined by the following two data and two axioms:

##\mathscr{C}## assigns to each object ##K## of ##\mathscr{K}## a class ##\mathscr{C}(K)## of ##\mathscr{C-}####\textit{structures on}## ##K.## A ##\mathscr{C-}####\textit{structure}## is a pair ##(K,s)## with ##s\in \mathscr{C}(K).##

For each ordered pair ##(K,S;L,t)## of ##\mathscr{C-}\textit{structures,} \mathscr{C,}## assigns a subset ##\mathscr{C}(s,t)## of ##\mathscr{K}(K,L)## of ##\mathscr{C}-\textit{admissible } \mathscr{K-}\textit{morphisms from } (K,s) \textit{to } (L,t);## to denote that ##f:K\rightarrow L## is in ##\mathscr{C}(s,t)## we will write ##f(K,s)\rightarrow (L,t)## of ##f:s\rightarrow t## (if necessary, imposing additional decoration should more than one ##\mathscr{C}## be in the picture).

The two axioms are:

##\textit{Axiom of Composition.}## If ##f:s\rightarrow t## and ##g:t\rightarrow u## then ##f\circ g:t\rightarrow u.##

##\textit{Structure is Abstract.}## If ##f:K\rightarrow L## is an isomorphism in ##\mathscr{K}## then for all ##t\in \mathscr{C}(L)## there exists unique ##s\in \mathscr{C}(K)## such that ##f:s\rightarrow t## and ##f^{-1}:t\rightarrow s.## optimal; that is, specifically, if whenever ##(K',s')## is a ##\mathscr{C-}##structure and ##g:K\rightarrow K'## is a ##\mathscr{K-}##morphism such that (please see the following image)


##f_i\circ g## is admissible for all ##i## then ##g## is also admissible.


Sets with structure-manes.webp


(From Arbib and Manes):

A ##\textbf{category, C, of sets with structure}## is given by the following two data and two axioms:

##\textbf{C}## assigns to each set ##X## a set ##\textbf{C}(X)## of ##\textbf{C-structures on } X##. ##\textbf{C-structure,}## then, is a pair ##(X,s)## with ##s## in ##\textbf{C}(X).## For each pair of sets ##(X,Y),\textbf{C}## assigns a function

$$\textbf{C}(X)\times \textbf{C}(Y)\rightarrow\textbf{P}(Y^{X}):s,t\mapsto \textbf{C}(s,t)$$

where ##Y^{X}## is, recall, the set of functions from ##X## to ##Y## and ##\textbf{P}(Y^{X})## is the set of its subsets. We write "##f:(X,s)\rightarrow (Y,t)##" and say "##f## is ##\textbf{admissible}## in ##\textbf{C}## from ##s## to ##t##" just in case ##f## is in ##\textbf{C}(s,t).## The axioms are:

##\textbf{Admissible maps to composable:}## If ##f:(X,s)\rightarrow (Y,t)## and ##g:(Y,t)\rightarrow (Z,u)## then ##g\circ f:(X,s)\rightarrow (Z,u).##

##\textbf{Structure is abstract:}## If ##f:X\rightarrow Y## is a bijection (i.e. an isomorphism of sets) and if ##t## is in ##\textbf{C}(Y)## there exists a unique ##s## in ##\textbf{C}(X)## with ##f:(X,s)\rightarrow (Y,t)## and ##f^{-1}:(Y,t)\rightarrow (X,s).##

##\textit{Some mappings between metric spaces.}## Let ##(X,d), (Y,e)## be metric spaces. A function ##f:X\to Y## is an ##\textbf{isometry}## ##\textit{from}## ##(X,d)## ##\textit{into}## ##(Y,e)## if for all ##x_1,x_2\in X, e(fx_1,fx_2)=d(x_1,x_2),## that is, if ##f## preserves distances. An isometry is automatically one-to-one, for if ##x_1\neq x_2## then ##e(fx_1,fx_2)=d(x_1,x_2)\neq 0,## which implies ##fx_1\neq fx_2. f## is an ##\textbf{isomorphism}## ##(X,d)\to (Y,e)## if ##f## is an isometry and ##f## is onto. In this case ##f^{-1}## is also an isomorphism, since ##d(f^{-1}y_1, f^{-1}y_2)=e(ff^{-1}y_1, ff^{-1}y_2)=e(y_1,y_2).##
##f:(X,d)\to (Y,e)## is a ##\textit{Lipschitz map}## if there exists ##\lambda >0## such that ##e(fx_1,fx_2)\leq \lambda d(x_1,x_2)## for all ##x_1,x_2\in X.##

##\textit{The categories}## ##\textbf{Met1}## ##\textit{and}## ##\textbf{Met*}.## Let ##\textbf{Met1}## be the category whose objects are metric spaces ##(X,d)## ##\textit{of diameter}\leq 1## (that is, ##d(x,y)\leq 1## for all ##x,y##) and whose morphisms are contractions. Let ##\textbf{Met*}## be the category whose objects are ##\textit{metric spaces with base point}##
##(X,d,x)## (meaning ##(X,d)## is a metric space and ##x## "the base point" is an arbitrary element of ##X##) and whose morphisms ##f:(X,d,x)\to (Y,e,y)## are contractions ##f:(X,d)\to (Y,e)## such that ##fx=y.## It is clear that ##\textbf{Met*}## is a category since if ##fx=y## and ##gy=z## then ##gfx=gy=z.##

##\textit{Every}## family in ##\textbf{Met*}## has a product, although it need not be built on the cartesian product set! Given ##(X_i,d_i,x_i)## let ##X## be the subset of all ##I-##tuples ##(a_i)## with each ##a_i\in X## such that the ##I-##tuple of numbers ##d_i(a_i,x_i)## has an upper bound, and define ##d((a_i),(b_i))=\text{Sup }\{d_i(a_i,b_i)\mid i\in I\}.## Then ##x=(x_i)## is in ##X.## Since there exist, by definition of ##X,## numbers ##s## and ##t## such that ##d_i(a_i,x_i)\leq t## for all ##i## and ##d_i(b_i,x_i)\leq s## for all ##i## then $$d_i(a_i,b_i)\leq d_i(a_i,x_i)+d_i(x,b_i)=d_i(a_i,x_i)+d_i(b_i,x_i)\leq t+s$$

which proves that ##d((a_i),(b_i))## is a well -defined number. Let ##p_i:(X,d_,x)\to (X_i,d_i,x_i)## be the usual coordinate projections.


##\color{green}{Exercise:}## Consider ##(\textbf{R}_+,|x-y|,0)## in ##\textbf{Met*}.## Show that, if ##X=\textbf{R}^I## with ##I## infinite, there is no structure ##(d,x)## in ##\textbf{Met*}(X)## such that each ##\pi_i:(X,d,x)\to (\textbf{R}_+, |x-y|,0)## is admissible.




Thank you in advance.
 
A sphere as topological manifold can be defined by gluing together the boundary of two disk. Basically one starts assigning each disk the subspace topology from ##\mathbb R^2## and then taking the quotient topology obtained by gluing their boundaries. Starting from the above definition of 2-sphere as topological manifold, shows that it is homeomorphic to the "embedded" sphere understood as subset of ##\mathbb R^3## in the subspace topology.

Similar threads

Replies
0
Views
388
Replies
11
Views
2K
3
Replies
100
Views
11K
Replies
2
Views
3K
4
Replies
175
Views
25K
Back
Top