Density of Baire’s Theorem for Complete Metric Spaces

The name of the pictureThe name of the pictureThe name of the pictureClash Royale CLAN TAG#URR8PPP











up vote
1
down vote

favorite













Let $G_n$ be a sequence of dense open subsets of a complete metric space $X$. Show that $ bigcap_n geq 1G_n$ is nonempty.




Proof:



Consider some point $p_1 in G_1$. There exist real numbers $0< r_1 <r<1$ such that
$$N_1 = N_r_1 (p_1) subset N_r (p_1) subset G_1.$$



Let $q$ be a limit point of $N_1$. For any $epsilon > 0$ we can find $x in X$ such that $x in N_1$ and $x in N_epsilon (q)$. Hence
$$d(p_1, q) leq d(p_1, x) + d(x, q)<r_1 + epsilon.$$
Since $epsilon$ was arbitrary it follows that
$$d(p_1, q) leq r_1.$$



Thus, if $q$ is any limit point of $N_1$, $q$ is an interior point of $G_1$ (to see this, simply consider the neighbourhood of radius $0<epsilon<r-r_1$) and so we can write



$overlineN_1 subset G_1.$



Since the $G_n$ are dense, either $p_1$ is a point in $G_2$ or a limit point of $G_2$. In either case, arguing as above, we can find a point $p_2 in G_2 $ and a neighbourhood of $p_2$ with radius $r_2 < minbig(r_1, frac12big)$ such that
$$overlineN_2 subset overlineN_1$$
and, since the $G_n$ are all open, we also choose $r_2$ small enough that
$$overlineN_2 subset G_2.$$



Continuing this process, we can construct the sequence of closed subsets $BigoverlineN_nBig$. Further, by the way we have chosen the $r_n$ (i.e., $0<r_n<frac1n$), it follows $r_n to 0$, so we have actually constructed a nested sequence of closed bounded sets $BigoverlineN_nBig$ such that $diam overlineN_n to 0.$ Since $X$ is a complete metric space, the intersection of this sequence is nonempty and since each $$overlineN_n subset G_n$$ the result follows. $qquad square$



Assuming this proof is correct, I am wondering if it can be extended to show that the intersection of the $G_n$ is actually dense, or if that requires a completely different line of argument?







share|cite|improve this question






















  • If $X$ is the empty space then each $G_n$ is empty and so is $cap_nG_n.$ This is a trivial case but sometimes an overlooked case can bite you .
    – DanielWainfleet
    Aug 23 at 4:39














up vote
1
down vote

favorite













Let $G_n$ be a sequence of dense open subsets of a complete metric space $X$. Show that $ bigcap_n geq 1G_n$ is nonempty.




Proof:



Consider some point $p_1 in G_1$. There exist real numbers $0< r_1 <r<1$ such that
$$N_1 = N_r_1 (p_1) subset N_r (p_1) subset G_1.$$



Let $q$ be a limit point of $N_1$. For any $epsilon > 0$ we can find $x in X$ such that $x in N_1$ and $x in N_epsilon (q)$. Hence
$$d(p_1, q) leq d(p_1, x) + d(x, q)<r_1 + epsilon.$$
Since $epsilon$ was arbitrary it follows that
$$d(p_1, q) leq r_1.$$



Thus, if $q$ is any limit point of $N_1$, $q$ is an interior point of $G_1$ (to see this, simply consider the neighbourhood of radius $0<epsilon<r-r_1$) and so we can write



$overlineN_1 subset G_1.$



Since the $G_n$ are dense, either $p_1$ is a point in $G_2$ or a limit point of $G_2$. In either case, arguing as above, we can find a point $p_2 in G_2 $ and a neighbourhood of $p_2$ with radius $r_2 < minbig(r_1, frac12big)$ such that
$$overlineN_2 subset overlineN_1$$
and, since the $G_n$ are all open, we also choose $r_2$ small enough that
$$overlineN_2 subset G_2.$$



Continuing this process, we can construct the sequence of closed subsets $BigoverlineN_nBig$. Further, by the way we have chosen the $r_n$ (i.e., $0<r_n<frac1n$), it follows $r_n to 0$, so we have actually constructed a nested sequence of closed bounded sets $BigoverlineN_nBig$ such that $diam overlineN_n to 0.$ Since $X$ is a complete metric space, the intersection of this sequence is nonempty and since each $$overlineN_n subset G_n$$ the result follows. $qquad square$



Assuming this proof is correct, I am wondering if it can be extended to show that the intersection of the $G_n$ is actually dense, or if that requires a completely different line of argument?







share|cite|improve this question






















  • If $X$ is the empty space then each $G_n$ is empty and so is $cap_nG_n.$ This is a trivial case but sometimes an overlooked case can bite you .
    – DanielWainfleet
    Aug 23 at 4:39












up vote
1
down vote

favorite









up vote
1
down vote

favorite












Let $G_n$ be a sequence of dense open subsets of a complete metric space $X$. Show that $ bigcap_n geq 1G_n$ is nonempty.




Proof:



Consider some point $p_1 in G_1$. There exist real numbers $0< r_1 <r<1$ such that
$$N_1 = N_r_1 (p_1) subset N_r (p_1) subset G_1.$$



Let $q$ be a limit point of $N_1$. For any $epsilon > 0$ we can find $x in X$ such that $x in N_1$ and $x in N_epsilon (q)$. Hence
$$d(p_1, q) leq d(p_1, x) + d(x, q)<r_1 + epsilon.$$
Since $epsilon$ was arbitrary it follows that
$$d(p_1, q) leq r_1.$$



Thus, if $q$ is any limit point of $N_1$, $q$ is an interior point of $G_1$ (to see this, simply consider the neighbourhood of radius $0<epsilon<r-r_1$) and so we can write



$overlineN_1 subset G_1.$



Since the $G_n$ are dense, either $p_1$ is a point in $G_2$ or a limit point of $G_2$. In either case, arguing as above, we can find a point $p_2 in G_2 $ and a neighbourhood of $p_2$ with radius $r_2 < minbig(r_1, frac12big)$ such that
$$overlineN_2 subset overlineN_1$$
and, since the $G_n$ are all open, we also choose $r_2$ small enough that
$$overlineN_2 subset G_2.$$



Continuing this process, we can construct the sequence of closed subsets $BigoverlineN_nBig$. Further, by the way we have chosen the $r_n$ (i.e., $0<r_n<frac1n$), it follows $r_n to 0$, so we have actually constructed a nested sequence of closed bounded sets $BigoverlineN_nBig$ such that $diam overlineN_n to 0.$ Since $X$ is a complete metric space, the intersection of this sequence is nonempty and since each $$overlineN_n subset G_n$$ the result follows. $qquad square$



Assuming this proof is correct, I am wondering if it can be extended to show that the intersection of the $G_n$ is actually dense, or if that requires a completely different line of argument?







share|cite|improve this question















Let $G_n$ be a sequence of dense open subsets of a complete metric space $X$. Show that $ bigcap_n geq 1G_n$ is nonempty.




Proof:



Consider some point $p_1 in G_1$. There exist real numbers $0< r_1 <r<1$ such that
$$N_1 = N_r_1 (p_1) subset N_r (p_1) subset G_1.$$



Let $q$ be a limit point of $N_1$. For any $epsilon > 0$ we can find $x in X$ such that $x in N_1$ and $x in N_epsilon (q)$. Hence
$$d(p_1, q) leq d(p_1, x) + d(x, q)<r_1 + epsilon.$$
Since $epsilon$ was arbitrary it follows that
$$d(p_1, q) leq r_1.$$



Thus, if $q$ is any limit point of $N_1$, $q$ is an interior point of $G_1$ (to see this, simply consider the neighbourhood of radius $0<epsilon<r-r_1$) and so we can write



$overlineN_1 subset G_1.$



Since the $G_n$ are dense, either $p_1$ is a point in $G_2$ or a limit point of $G_2$. In either case, arguing as above, we can find a point $p_2 in G_2 $ and a neighbourhood of $p_2$ with radius $r_2 < minbig(r_1, frac12big)$ such that
$$overlineN_2 subset overlineN_1$$
and, since the $G_n$ are all open, we also choose $r_2$ small enough that
$$overlineN_2 subset G_2.$$



Continuing this process, we can construct the sequence of closed subsets $BigoverlineN_nBig$. Further, by the way we have chosen the $r_n$ (i.e., $0<r_n<frac1n$), it follows $r_n to 0$, so we have actually constructed a nested sequence of closed bounded sets $BigoverlineN_nBig$ such that $diam overlineN_n to 0.$ Since $X$ is a complete metric space, the intersection of this sequence is nonempty and since each $$overlineN_n subset G_n$$ the result follows. $qquad square$



Assuming this proof is correct, I am wondering if it can be extended to show that the intersection of the $G_n$ is actually dense, or if that requires a completely different line of argument?









share|cite|improve this question













share|cite|improve this question




share|cite|improve this question








edited Jul 7 at 3:44

























asked Jul 6 at 4:20









Moed Pol Bollo

28228




28228











  • If $X$ is the empty space then each $G_n$ is empty and so is $cap_nG_n.$ This is a trivial case but sometimes an overlooked case can bite you .
    – DanielWainfleet
    Aug 23 at 4:39
















  • If $X$ is the empty space then each $G_n$ is empty and so is $cap_nG_n.$ This is a trivial case but sometimes an overlooked case can bite you .
    – DanielWainfleet
    Aug 23 at 4:39















If $X$ is the empty space then each $G_n$ is empty and so is $cap_nG_n.$ This is a trivial case but sometimes an overlooked case can bite you .
– DanielWainfleet
Aug 23 at 4:39




If $X$ is the empty space then each $G_n$ is empty and so is $cap_nG_n.$ This is a trivial case but sometimes an overlooked case can bite you .
– DanielWainfleet
Aug 23 at 4:39










2 Answers
2






active

oldest

votes

















up vote
3
down vote



accepted










That the intersection of the $G_n$ is dense, is a small modification of the above argument: let $O$ be any non-empty open set in $X$, and start with an open ball $N_0 subseteq O$ and stay inside $N_0$ with all subsequent steps. This hardly takes any effort at all, but does show that the $x in cap G_n$ is also in $N_0$ hence in $O$. So the intersection of the $G_n$ intersects every non-empty open set, hence is dense.



The construction of the $N_n$ can be a bit simplified:



  • Start with $p_0 in O$ and $N_0 := B(p, r_0) subseteq O$.

  • $N_0$ is open, so $G_1 cap N_0$ is non-empty (as $G_1$ is dense) and open (as $G_1$ is open). So pick $p_1 in G_1 cap N_0$ and let $0< r_1 < 1$ be small enough that $overlineB(p_1, r_1) subseteq G_1 cap N_0$. Define $N_1 = B(p_1, r_1)$.

  • $N_1$ is open and again we have that $p_2 in N_1 cap G_0 cap G_1$ exists by denseness and this set is open so there exists $0<r_2 < frac12$ such that $overlineB(p_2, r_2) subseteq (N_1 cap G_0 cap G_1)$. Define $N_2 = B(p_2, r_2)$.

  • continue this process recursively.

No distinguishing limit points etc. Just go straight to the goal.



Then the $overlineN_n$ $n ge 1$ form the required nested family that the Cantor intersection theorem can be applied to. The promised $p in bigcap_n ge 1 overlineN_n subseteq O cap bigcap_n ge 1O_n$ witnesses the denseness of $bigcap_n ge 1 G_n$.






share|cite|improve this answer




















  • Quick question, though: how can you be sure that $overlineB(p_1,r_1) subseteq G_1 cap N_0$, without reference to limit points?
    – Moed Pol Bollo
    Jul 8 at 1:45










  • @MoedPolBollo Because $N_0 cap G_1$ is open and non-empty, as the intersection of an open and a dense open set. We pick $p_1$ in it. This is an interior point of $N_0 cap G_1$, so some ball around it sits inside $N_0 cap G_1$, say $B(p,s_1) subseteq N_0 cap G_1$, for some $s_1 >0$. Then take $r_1 = min(1, fracs_12)$ and we have $overlineB(p_1,r_1) subseteq B(p, s_1) subseteq N_0 cap G_1$ etc. In any metric space, for an open balls, if $r <s$ we know that $overlineB(p,r) subseteq x: d(p,x) le r subseteq B(p,s)$. It's easy to get smaller closures this way.
    – Henno Brandsma
    Jul 8 at 5:49










  • To the proposer: In any metric space$ (X,d)$ and any $p_1in X,$ if $0<r_1<r$ then $overline N_r_1(p_1)subset N_r(p_1)$ because we can easily show that $overline N_r_1(p_1)subset q:d(q,p_1)leq r_1.$ ... (That is, if $d(p,q)=r_1+s>r_1$ then by the triangle inequality, the open ball $N_s/2(q) $ is disjoint from $N_r_1(p)$....).... So from the first displayed line of your proof we have $overline N_1subset N_r(p_1)subset G_1.$
    – DanielWainfleet
    Aug 23 at 4:48


















up vote
-1
down vote













The proofs of @Bollo and @Brandsma are great and I try to supplement a situation when the metric space is finite.



Supplement proof:



When the metric space $X$ is a finite set, it is always complete. However, there is no such open set $N_r(p_1) subset G_1$ (in your proof) because every point is an isolated point. Therefore the only dense open subset of $X$ is itself. So the intersection is itself, which is dense.






share|cite|improve this answer




















    Your Answer




    StackExchange.ifUsing("editor", function ()
    return StackExchange.using("mathjaxEditing", function ()
    StackExchange.MarkdownEditor.creationCallbacks.add(function (editor, postfix)
    StackExchange.mathjaxEditing.prepareWmdForMathJax(editor, postfix, [["$", "$"], ["\\(","\\)"]]);
    );
    );
    , "mathjax-editing");

    StackExchange.ready(function()
    var channelOptions =
    tags: "".split(" "),
    id: "69"
    ;
    initTagRenderer("".split(" "), "".split(" "), channelOptions);

    StackExchange.using("externalEditor", function()
    // Have to fire editor after snippets, if snippets enabled
    if (StackExchange.settings.snippets.snippetsEnabled)
    StackExchange.using("snippets", function()
    createEditor();
    );

    else
    createEditor();

    );

    function createEditor()
    StackExchange.prepareEditor(
    heartbeatType: 'answer',
    convertImagesToLinks: true,
    noModals: false,
    showLowRepImageUploadWarning: true,
    reputationToPostImages: 10,
    bindNavPrevention: true,
    postfix: "",
    noCode: true, onDemand: true,
    discardSelector: ".discard-answer"
    ,immediatelyShowMarkdownHelp:true
    );



    );













     

    draft saved


    draft discarded


















    StackExchange.ready(
    function ()
    StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fmath.stackexchange.com%2fquestions%2f2842488%2fdensity-of-baire-s-theorem-for-complete-metric-spaces%23new-answer', 'question_page');

    );

    Post as a guest






























    2 Answers
    2






    active

    oldest

    votes








    2 Answers
    2






    active

    oldest

    votes









    active

    oldest

    votes






    active

    oldest

    votes








    up vote
    3
    down vote



    accepted










    That the intersection of the $G_n$ is dense, is a small modification of the above argument: let $O$ be any non-empty open set in $X$, and start with an open ball $N_0 subseteq O$ and stay inside $N_0$ with all subsequent steps. This hardly takes any effort at all, but does show that the $x in cap G_n$ is also in $N_0$ hence in $O$. So the intersection of the $G_n$ intersects every non-empty open set, hence is dense.



    The construction of the $N_n$ can be a bit simplified:



    • Start with $p_0 in O$ and $N_0 := B(p, r_0) subseteq O$.

    • $N_0$ is open, so $G_1 cap N_0$ is non-empty (as $G_1$ is dense) and open (as $G_1$ is open). So pick $p_1 in G_1 cap N_0$ and let $0< r_1 < 1$ be small enough that $overlineB(p_1, r_1) subseteq G_1 cap N_0$. Define $N_1 = B(p_1, r_1)$.

    • $N_1$ is open and again we have that $p_2 in N_1 cap G_0 cap G_1$ exists by denseness and this set is open so there exists $0<r_2 < frac12$ such that $overlineB(p_2, r_2) subseteq (N_1 cap G_0 cap G_1)$. Define $N_2 = B(p_2, r_2)$.

    • continue this process recursively.

    No distinguishing limit points etc. Just go straight to the goal.



    Then the $overlineN_n$ $n ge 1$ form the required nested family that the Cantor intersection theorem can be applied to. The promised $p in bigcap_n ge 1 overlineN_n subseteq O cap bigcap_n ge 1O_n$ witnesses the denseness of $bigcap_n ge 1 G_n$.






    share|cite|improve this answer




















    • Quick question, though: how can you be sure that $overlineB(p_1,r_1) subseteq G_1 cap N_0$, without reference to limit points?
      – Moed Pol Bollo
      Jul 8 at 1:45










    • @MoedPolBollo Because $N_0 cap G_1$ is open and non-empty, as the intersection of an open and a dense open set. We pick $p_1$ in it. This is an interior point of $N_0 cap G_1$, so some ball around it sits inside $N_0 cap G_1$, say $B(p,s_1) subseteq N_0 cap G_1$, for some $s_1 >0$. Then take $r_1 = min(1, fracs_12)$ and we have $overlineB(p_1,r_1) subseteq B(p, s_1) subseteq N_0 cap G_1$ etc. In any metric space, for an open balls, if $r <s$ we know that $overlineB(p,r) subseteq x: d(p,x) le r subseteq B(p,s)$. It's easy to get smaller closures this way.
      – Henno Brandsma
      Jul 8 at 5:49










    • To the proposer: In any metric space$ (X,d)$ and any $p_1in X,$ if $0<r_1<r$ then $overline N_r_1(p_1)subset N_r(p_1)$ because we can easily show that $overline N_r_1(p_1)subset q:d(q,p_1)leq r_1.$ ... (That is, if $d(p,q)=r_1+s>r_1$ then by the triangle inequality, the open ball $N_s/2(q) $ is disjoint from $N_r_1(p)$....).... So from the first displayed line of your proof we have $overline N_1subset N_r(p_1)subset G_1.$
      – DanielWainfleet
      Aug 23 at 4:48















    up vote
    3
    down vote



    accepted










    That the intersection of the $G_n$ is dense, is a small modification of the above argument: let $O$ be any non-empty open set in $X$, and start with an open ball $N_0 subseteq O$ and stay inside $N_0$ with all subsequent steps. This hardly takes any effort at all, but does show that the $x in cap G_n$ is also in $N_0$ hence in $O$. So the intersection of the $G_n$ intersects every non-empty open set, hence is dense.



    The construction of the $N_n$ can be a bit simplified:



    • Start with $p_0 in O$ and $N_0 := B(p, r_0) subseteq O$.

    • $N_0$ is open, so $G_1 cap N_0$ is non-empty (as $G_1$ is dense) and open (as $G_1$ is open). So pick $p_1 in G_1 cap N_0$ and let $0< r_1 < 1$ be small enough that $overlineB(p_1, r_1) subseteq G_1 cap N_0$. Define $N_1 = B(p_1, r_1)$.

    • $N_1$ is open and again we have that $p_2 in N_1 cap G_0 cap G_1$ exists by denseness and this set is open so there exists $0<r_2 < frac12$ such that $overlineB(p_2, r_2) subseteq (N_1 cap G_0 cap G_1)$. Define $N_2 = B(p_2, r_2)$.

    • continue this process recursively.

    No distinguishing limit points etc. Just go straight to the goal.



    Then the $overlineN_n$ $n ge 1$ form the required nested family that the Cantor intersection theorem can be applied to. The promised $p in bigcap_n ge 1 overlineN_n subseteq O cap bigcap_n ge 1O_n$ witnesses the denseness of $bigcap_n ge 1 G_n$.






    share|cite|improve this answer




















    • Quick question, though: how can you be sure that $overlineB(p_1,r_1) subseteq G_1 cap N_0$, without reference to limit points?
      – Moed Pol Bollo
      Jul 8 at 1:45










    • @MoedPolBollo Because $N_0 cap G_1$ is open and non-empty, as the intersection of an open and a dense open set. We pick $p_1$ in it. This is an interior point of $N_0 cap G_1$, so some ball around it sits inside $N_0 cap G_1$, say $B(p,s_1) subseteq N_0 cap G_1$, for some $s_1 >0$. Then take $r_1 = min(1, fracs_12)$ and we have $overlineB(p_1,r_1) subseteq B(p, s_1) subseteq N_0 cap G_1$ etc. In any metric space, for an open balls, if $r <s$ we know that $overlineB(p,r) subseteq x: d(p,x) le r subseteq B(p,s)$. It's easy to get smaller closures this way.
      – Henno Brandsma
      Jul 8 at 5:49










    • To the proposer: In any metric space$ (X,d)$ and any $p_1in X,$ if $0<r_1<r$ then $overline N_r_1(p_1)subset N_r(p_1)$ because we can easily show that $overline N_r_1(p_1)subset q:d(q,p_1)leq r_1.$ ... (That is, if $d(p,q)=r_1+s>r_1$ then by the triangle inequality, the open ball $N_s/2(q) $ is disjoint from $N_r_1(p)$....).... So from the first displayed line of your proof we have $overline N_1subset N_r(p_1)subset G_1.$
      – DanielWainfleet
      Aug 23 at 4:48













    up vote
    3
    down vote



    accepted







    up vote
    3
    down vote



    accepted






    That the intersection of the $G_n$ is dense, is a small modification of the above argument: let $O$ be any non-empty open set in $X$, and start with an open ball $N_0 subseteq O$ and stay inside $N_0$ with all subsequent steps. This hardly takes any effort at all, but does show that the $x in cap G_n$ is also in $N_0$ hence in $O$. So the intersection of the $G_n$ intersects every non-empty open set, hence is dense.



    The construction of the $N_n$ can be a bit simplified:



    • Start with $p_0 in O$ and $N_0 := B(p, r_0) subseteq O$.

    • $N_0$ is open, so $G_1 cap N_0$ is non-empty (as $G_1$ is dense) and open (as $G_1$ is open). So pick $p_1 in G_1 cap N_0$ and let $0< r_1 < 1$ be small enough that $overlineB(p_1, r_1) subseteq G_1 cap N_0$. Define $N_1 = B(p_1, r_1)$.

    • $N_1$ is open and again we have that $p_2 in N_1 cap G_0 cap G_1$ exists by denseness and this set is open so there exists $0<r_2 < frac12$ such that $overlineB(p_2, r_2) subseteq (N_1 cap G_0 cap G_1)$. Define $N_2 = B(p_2, r_2)$.

    • continue this process recursively.

    No distinguishing limit points etc. Just go straight to the goal.



    Then the $overlineN_n$ $n ge 1$ form the required nested family that the Cantor intersection theorem can be applied to. The promised $p in bigcap_n ge 1 overlineN_n subseteq O cap bigcap_n ge 1O_n$ witnesses the denseness of $bigcap_n ge 1 G_n$.






    share|cite|improve this answer












    That the intersection of the $G_n$ is dense, is a small modification of the above argument: let $O$ be any non-empty open set in $X$, and start with an open ball $N_0 subseteq O$ and stay inside $N_0$ with all subsequent steps. This hardly takes any effort at all, but does show that the $x in cap G_n$ is also in $N_0$ hence in $O$. So the intersection of the $G_n$ intersects every non-empty open set, hence is dense.



    The construction of the $N_n$ can be a bit simplified:



    • Start with $p_0 in O$ and $N_0 := B(p, r_0) subseteq O$.

    • $N_0$ is open, so $G_1 cap N_0$ is non-empty (as $G_1$ is dense) and open (as $G_1$ is open). So pick $p_1 in G_1 cap N_0$ and let $0< r_1 < 1$ be small enough that $overlineB(p_1, r_1) subseteq G_1 cap N_0$. Define $N_1 = B(p_1, r_1)$.

    • $N_1$ is open and again we have that $p_2 in N_1 cap G_0 cap G_1$ exists by denseness and this set is open so there exists $0<r_2 < frac12$ such that $overlineB(p_2, r_2) subseteq (N_1 cap G_0 cap G_1)$. Define $N_2 = B(p_2, r_2)$.

    • continue this process recursively.

    No distinguishing limit points etc. Just go straight to the goal.



    Then the $overlineN_n$ $n ge 1$ form the required nested family that the Cantor intersection theorem can be applied to. The promised $p in bigcap_n ge 1 overlineN_n subseteq O cap bigcap_n ge 1O_n$ witnesses the denseness of $bigcap_n ge 1 G_n$.







    share|cite|improve this answer












    share|cite|improve this answer



    share|cite|improve this answer










    answered Jul 7 at 17:56









    Henno Brandsma

    92.5k342100




    92.5k342100











    • Quick question, though: how can you be sure that $overlineB(p_1,r_1) subseteq G_1 cap N_0$, without reference to limit points?
      – Moed Pol Bollo
      Jul 8 at 1:45










    • @MoedPolBollo Because $N_0 cap G_1$ is open and non-empty, as the intersection of an open and a dense open set. We pick $p_1$ in it. This is an interior point of $N_0 cap G_1$, so some ball around it sits inside $N_0 cap G_1$, say $B(p,s_1) subseteq N_0 cap G_1$, for some $s_1 >0$. Then take $r_1 = min(1, fracs_12)$ and we have $overlineB(p_1,r_1) subseteq B(p, s_1) subseteq N_0 cap G_1$ etc. In any metric space, for an open balls, if $r <s$ we know that $overlineB(p,r) subseteq x: d(p,x) le r subseteq B(p,s)$. It's easy to get smaller closures this way.
      – Henno Brandsma
      Jul 8 at 5:49










    • To the proposer: In any metric space$ (X,d)$ and any $p_1in X,$ if $0<r_1<r$ then $overline N_r_1(p_1)subset N_r(p_1)$ because we can easily show that $overline N_r_1(p_1)subset q:d(q,p_1)leq r_1.$ ... (That is, if $d(p,q)=r_1+s>r_1$ then by the triangle inequality, the open ball $N_s/2(q) $ is disjoint from $N_r_1(p)$....).... So from the first displayed line of your proof we have $overline N_1subset N_r(p_1)subset G_1.$
      – DanielWainfleet
      Aug 23 at 4:48

















    • Quick question, though: how can you be sure that $overlineB(p_1,r_1) subseteq G_1 cap N_0$, without reference to limit points?
      – Moed Pol Bollo
      Jul 8 at 1:45










    • @MoedPolBollo Because $N_0 cap G_1$ is open and non-empty, as the intersection of an open and a dense open set. We pick $p_1$ in it. This is an interior point of $N_0 cap G_1$, so some ball around it sits inside $N_0 cap G_1$, say $B(p,s_1) subseteq N_0 cap G_1$, for some $s_1 >0$. Then take $r_1 = min(1, fracs_12)$ and we have $overlineB(p_1,r_1) subseteq B(p, s_1) subseteq N_0 cap G_1$ etc. In any metric space, for an open balls, if $r <s$ we know that $overlineB(p,r) subseteq x: d(p,x) le r subseteq B(p,s)$. It's easy to get smaller closures this way.
      – Henno Brandsma
      Jul 8 at 5:49










    • To the proposer: In any metric space$ (X,d)$ and any $p_1in X,$ if $0<r_1<r$ then $overline N_r_1(p_1)subset N_r(p_1)$ because we can easily show that $overline N_r_1(p_1)subset q:d(q,p_1)leq r_1.$ ... (That is, if $d(p,q)=r_1+s>r_1$ then by the triangle inequality, the open ball $N_s/2(q) $ is disjoint from $N_r_1(p)$....).... So from the first displayed line of your proof we have $overline N_1subset N_r(p_1)subset G_1.$
      – DanielWainfleet
      Aug 23 at 4:48
















    Quick question, though: how can you be sure that $overlineB(p_1,r_1) subseteq G_1 cap N_0$, without reference to limit points?
    – Moed Pol Bollo
    Jul 8 at 1:45




    Quick question, though: how can you be sure that $overlineB(p_1,r_1) subseteq G_1 cap N_0$, without reference to limit points?
    – Moed Pol Bollo
    Jul 8 at 1:45












    @MoedPolBollo Because $N_0 cap G_1$ is open and non-empty, as the intersection of an open and a dense open set. We pick $p_1$ in it. This is an interior point of $N_0 cap G_1$, so some ball around it sits inside $N_0 cap G_1$, say $B(p,s_1) subseteq N_0 cap G_1$, for some $s_1 >0$. Then take $r_1 = min(1, fracs_12)$ and we have $overlineB(p_1,r_1) subseteq B(p, s_1) subseteq N_0 cap G_1$ etc. In any metric space, for an open balls, if $r <s$ we know that $overlineB(p,r) subseteq x: d(p,x) le r subseteq B(p,s)$. It's easy to get smaller closures this way.
    – Henno Brandsma
    Jul 8 at 5:49




    @MoedPolBollo Because $N_0 cap G_1$ is open and non-empty, as the intersection of an open and a dense open set. We pick $p_1$ in it. This is an interior point of $N_0 cap G_1$, so some ball around it sits inside $N_0 cap G_1$, say $B(p,s_1) subseteq N_0 cap G_1$, for some $s_1 >0$. Then take $r_1 = min(1, fracs_12)$ and we have $overlineB(p_1,r_1) subseteq B(p, s_1) subseteq N_0 cap G_1$ etc. In any metric space, for an open balls, if $r <s$ we know that $overlineB(p,r) subseteq x: d(p,x) le r subseteq B(p,s)$. It's easy to get smaller closures this way.
    – Henno Brandsma
    Jul 8 at 5:49












    To the proposer: In any metric space$ (X,d)$ and any $p_1in X,$ if $0<r_1<r$ then $overline N_r_1(p_1)subset N_r(p_1)$ because we can easily show that $overline N_r_1(p_1)subset q:d(q,p_1)leq r_1.$ ... (That is, if $d(p,q)=r_1+s>r_1$ then by the triangle inequality, the open ball $N_s/2(q) $ is disjoint from $N_r_1(p)$....).... So from the first displayed line of your proof we have $overline N_1subset N_r(p_1)subset G_1.$
    – DanielWainfleet
    Aug 23 at 4:48





    To the proposer: In any metric space$ (X,d)$ and any $p_1in X,$ if $0<r_1<r$ then $overline N_r_1(p_1)subset N_r(p_1)$ because we can easily show that $overline N_r_1(p_1)subset q:d(q,p_1)leq r_1.$ ... (That is, if $d(p,q)=r_1+s>r_1$ then by the triangle inequality, the open ball $N_s/2(q) $ is disjoint from $N_r_1(p)$....).... So from the first displayed line of your proof we have $overline N_1subset N_r(p_1)subset G_1.$
    – DanielWainfleet
    Aug 23 at 4:48











    up vote
    -1
    down vote













    The proofs of @Bollo and @Brandsma are great and I try to supplement a situation when the metric space is finite.



    Supplement proof:



    When the metric space $X$ is a finite set, it is always complete. However, there is no such open set $N_r(p_1) subset G_1$ (in your proof) because every point is an isolated point. Therefore the only dense open subset of $X$ is itself. So the intersection is itself, which is dense.






    share|cite|improve this answer
























      up vote
      -1
      down vote













      The proofs of @Bollo and @Brandsma are great and I try to supplement a situation when the metric space is finite.



      Supplement proof:



      When the metric space $X$ is a finite set, it is always complete. However, there is no such open set $N_r(p_1) subset G_1$ (in your proof) because every point is an isolated point. Therefore the only dense open subset of $X$ is itself. So the intersection is itself, which is dense.






      share|cite|improve this answer






















        up vote
        -1
        down vote










        up vote
        -1
        down vote









        The proofs of @Bollo and @Brandsma are great and I try to supplement a situation when the metric space is finite.



        Supplement proof:



        When the metric space $X$ is a finite set, it is always complete. However, there is no such open set $N_r(p_1) subset G_1$ (in your proof) because every point is an isolated point. Therefore the only dense open subset of $X$ is itself. So the intersection is itself, which is dense.






        share|cite|improve this answer












        The proofs of @Bollo and @Brandsma are great and I try to supplement a situation when the metric space is finite.



        Supplement proof:



        When the metric space $X$ is a finite set, it is always complete. However, there is no such open set $N_r(p_1) subset G_1$ (in your proof) because every point is an isolated point. Therefore the only dense open subset of $X$ is itself. So the intersection is itself, which is dense.







        share|cite|improve this answer












        share|cite|improve this answer



        share|cite|improve this answer










        answered Aug 23 at 3:03









        Tengerye

        407




        407



























             

            draft saved


            draft discarded















































             


            draft saved


            draft discarded














            StackExchange.ready(
            function ()
            StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fmath.stackexchange.com%2fquestions%2f2842488%2fdensity-of-baire-s-theorem-for-complete-metric-spaces%23new-answer', 'question_page');

            );

            Post as a guest













































































            這個網誌中的熱門文章

            tkz-euclide: tkzDrawCircle[R] not working

            How to combine Bézier curves to a surface?

            1st Magritte Awards