inquality $sum_i=1^nfracsqrtx_i-x_i-1x_ilesum_i=1^n^2frac1i-frac12$

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











up vote
8
down vote

favorite
3












let $x_iin N^+$,and such $1=x_0le x_1le x_2lecdotsle x_n$ show that
$$sum_i=1^ndfracsqrtx_i-x_i-1x_ilesum_i=1^n^2dfrac1i-dfrac12$$



maybe can use C-S,But I can't it










share|cite|improve this question























  • What is wrong with my answer?
    – Francisco Abusleme
    Sep 6 at 18:49














up vote
8
down vote

favorite
3












let $x_iin N^+$,and such $1=x_0le x_1le x_2lecdotsle x_n$ show that
$$sum_i=1^ndfracsqrtx_i-x_i-1x_ilesum_i=1^n^2dfrac1i-dfrac12$$



maybe can use C-S,But I can't it










share|cite|improve this question























  • What is wrong with my answer?
    – Francisco Abusleme
    Sep 6 at 18:49












up vote
8
down vote

favorite
3









up vote
8
down vote

favorite
3






3





let $x_iin N^+$,and such $1=x_0le x_1le x_2lecdotsle x_n$ show that
$$sum_i=1^ndfracsqrtx_i-x_i-1x_ilesum_i=1^n^2dfrac1i-dfrac12$$



maybe can use C-S,But I can't it










share|cite|improve this question















let $x_iin N^+$,and such $1=x_0le x_1le x_2lecdotsle x_n$ show that
$$sum_i=1^ndfracsqrtx_i-x_i-1x_ilesum_i=1^n^2dfrac1i-dfrac12$$



maybe can use C-S,But I can't it







sequences-and-series inequality summation contest-math radicals






share|cite|improve this question















share|cite|improve this question













share|cite|improve this question




share|cite|improve this question








edited Sep 1 at 23:26









Arjang

5,48962162




5,48962162










asked Sep 1 at 4:15









communnites

1,2581431




1,2581431











  • What is wrong with my answer?
    – Francisco Abusleme
    Sep 6 at 18:49
















  • What is wrong with my answer?
    – Francisco Abusleme
    Sep 6 at 18:49















What is wrong with my answer?
– Francisco Abusleme
Sep 6 at 18:49




What is wrong with my answer?
– Francisco Abusleme
Sep 6 at 18:49










2 Answers
2






active

oldest

votes

















up vote
2
down vote













I think the following can help.



Let for $n>1$ we have $x_nleq n^2$.



Thus, $$sum_i=1^nfracsqrtx_i-x_i-1x_ileqsum_i=1^nfracx_i-x_i-1x_i=sum_i=1^nsum_j=1^x_i-x_i-1frac1x_ileq$$
$$leqsum_i=1^nsum_j=1^x_i-x_i-1frac1x_i-1+j=sum_i=x_0+1^x_nfrac1ileqsum_i=2^n^2frac1i<sum_i=1^n^2frac1i-frac12.$$
If first $x_k>n^2$.



Thus, $$sum_i=k^nfracsqrtx_i-x_i-1x_i<sum_i=k^nfrac1sqrtx_i<sum_i=k^nfrac1n<1$$ and
$$sum_i=1^k-1fracsqrtx_i-x_i-1x_i<sum_i=2^n^2frac1i.$$






share|cite|improve this answer






















  • then How to do it?
    – communnites
    Sep 1 at 8:35










  • @communnites For $x_nleq n^2$ it's proven. Try to end the proof if $x_k>n^2$ for some $k<n$, where $x_k-1leq n^2.$ The biggest part of the work I made already.
    – Michael Rozenberg
    Sep 1 at 8:51











  • why $sqrtx_i-x_i-1le x_i-x_i-1$? if $x_i=x_i-1$ then?
    – communnites
    Sep 1 at 8:57











  • and My question right is $sum_i=1^n^2dfrac1i^2-frac12$,not $sum_i=1^n^2dfrac1i^2$
    – communnites
    Sep 1 at 9:02











  • @communnites Because for all natural $m$ we have $sqrtmleq m.$ Also, for $x_nleq n^2$ I proved that $LHSleqsumlimits_i=2^n^2frac1i<sumlimits_i=1^n^2frac1i-frac12.$
    – Michael Rozenberg
    Sep 1 at 9:39


















up vote
-1
down vote













I am going to treat $x_i$ as a real variable. Then, we can differentiate the terms with respect to $x_i$:



$$fracddx_ifracsqrtx_i-x_i-1x_i$$



It is easy to show that the derivative is $ 0$ when $x_i=2x_i-1$, and it is also easy to show that it is a maximum (take the 2nd derivave, or just look at the cases $x_i=x_i-1$ and $x_irightarrowinfty $ .



So replacing $x_i=2x_i-1$ we get that each term becomes $frac12sqrtx_i-1$.



Note too, that when using this values for $x_i$, the smaller the values we choose, the term becomes bigger, so there is no doubt that the sum is maxed when we use $x_i=2^i$, therefore:



$$sum_i=1^n fracsqrtx_i-x_i-1x_i leqsum_i=0^n frac12sqrt2^i $$



So if show the worst case we are done. This last series can be easily solved as a geometric one, so the problem is now reduced to showing that:



$$(1+frac 1sqrt2)(1-(frac 1sqrt2)^n+1) leq sum_i=1^n^2 frac 1i - frac12 quad mathbf(1) $$



I will do it by induction. The case $n=1$ is easy. Now we take $(1)$ as our hypothesis, and we want to prove that if the case for $n$ is true, then the case for $n+1$ is true. Starting by the $n+1$ case in the right hand side, we have that:



$$sum_i=1^(n+1)^2 frac 1i - frac12 = sum_i=1^n^2 frac 1i - frac12 +sum_i=n^2+1^(n+1)^2 frac 1i $$



Now we use $(1)$ with the sum up to $n^2$ (and the negative one half), and since we can replace any quantity with a smaller one and the inequality will still be valid, I will replace the latter sum by the number of terms of the sum (expand $(n+1)^2$ to see that the number of terms is $2n+1$ ) multiplied by the smallest term, so now we have to show that:



$$ (1+frac 1sqrt2)(1-(frac 1sqrt2)^n+1)+ frac2n+1(n+1)^2geq (1+frac 1sqrt2)(1-(frac 1sqrt2)^n+2) $$



Rearranging:



$$frac2n+1(n+1)^2geq frac 12 (frac1sqrt2)^n+1 quad mathbf(2)$$



And here I will use a second induction. First, you show that the inequality holds for $n in text1,2,3,4,5$, and using $(2)$ as the hypothesis, we work the left hand side for the case $n+1$ as follows:



$$frac2n+3(n+2)^2 gt frac2n+3sqrt2(n+1)^2 $$



This last inequality, holds for $n gt 5$, because for these values of n, we have that $sqrt[4]2(n+1) gt n+2 $ (That's why I trated the other 5 cases separately). So now we can replace the hypothesis to get:



$$frac2n+3sqrt2(n+1)^2 = frac 12sqrt2 (frac 1sqrt2)^n+1 + frac22sqrt2(n+1)^2 gt frac 12 (frac 1sqrt2)^n+2 $$



Therefore,



$$frac2n+3(n+2)^2 gt frac 12 (frac 1sqrt2)^n+2 $$



This means the $n+1$ case is true, then the hypothesis $(2)$ is true for all natural $n$, and the same goes to the hypothesis $(1)$ and the proof is completed:



$$sum_i=1^n fracsqrtx_i-x_i-1x_i leqsum_i=0^n frac12sqrt2^i leq sum_i=1^n^2 frac 1i - frac12 $$






share|cite|improve this answer




















  • it's right? Thanks
    – communnites
    Sep 2 at 2:01










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%2f2901361%2finquality-sum-i-1n-frac-sqrtx-i-x-i-1x-i-le-sum-i-1n2-f%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
2
down vote













I think the following can help.



Let for $n>1$ we have $x_nleq n^2$.



Thus, $$sum_i=1^nfracsqrtx_i-x_i-1x_ileqsum_i=1^nfracx_i-x_i-1x_i=sum_i=1^nsum_j=1^x_i-x_i-1frac1x_ileq$$
$$leqsum_i=1^nsum_j=1^x_i-x_i-1frac1x_i-1+j=sum_i=x_0+1^x_nfrac1ileqsum_i=2^n^2frac1i<sum_i=1^n^2frac1i-frac12.$$
If first $x_k>n^2$.



Thus, $$sum_i=k^nfracsqrtx_i-x_i-1x_i<sum_i=k^nfrac1sqrtx_i<sum_i=k^nfrac1n<1$$ and
$$sum_i=1^k-1fracsqrtx_i-x_i-1x_i<sum_i=2^n^2frac1i.$$






share|cite|improve this answer






















  • then How to do it?
    – communnites
    Sep 1 at 8:35










  • @communnites For $x_nleq n^2$ it's proven. Try to end the proof if $x_k>n^2$ for some $k<n$, where $x_k-1leq n^2.$ The biggest part of the work I made already.
    – Michael Rozenberg
    Sep 1 at 8:51











  • why $sqrtx_i-x_i-1le x_i-x_i-1$? if $x_i=x_i-1$ then?
    – communnites
    Sep 1 at 8:57











  • and My question right is $sum_i=1^n^2dfrac1i^2-frac12$,not $sum_i=1^n^2dfrac1i^2$
    – communnites
    Sep 1 at 9:02











  • @communnites Because for all natural $m$ we have $sqrtmleq m.$ Also, for $x_nleq n^2$ I proved that $LHSleqsumlimits_i=2^n^2frac1i<sumlimits_i=1^n^2frac1i-frac12.$
    – Michael Rozenberg
    Sep 1 at 9:39















up vote
2
down vote













I think the following can help.



Let for $n>1$ we have $x_nleq n^2$.



Thus, $$sum_i=1^nfracsqrtx_i-x_i-1x_ileqsum_i=1^nfracx_i-x_i-1x_i=sum_i=1^nsum_j=1^x_i-x_i-1frac1x_ileq$$
$$leqsum_i=1^nsum_j=1^x_i-x_i-1frac1x_i-1+j=sum_i=x_0+1^x_nfrac1ileqsum_i=2^n^2frac1i<sum_i=1^n^2frac1i-frac12.$$
If first $x_k>n^2$.



Thus, $$sum_i=k^nfracsqrtx_i-x_i-1x_i<sum_i=k^nfrac1sqrtx_i<sum_i=k^nfrac1n<1$$ and
$$sum_i=1^k-1fracsqrtx_i-x_i-1x_i<sum_i=2^n^2frac1i.$$






share|cite|improve this answer






















  • then How to do it?
    – communnites
    Sep 1 at 8:35










  • @communnites For $x_nleq n^2$ it's proven. Try to end the proof if $x_k>n^2$ for some $k<n$, where $x_k-1leq n^2.$ The biggest part of the work I made already.
    – Michael Rozenberg
    Sep 1 at 8:51











  • why $sqrtx_i-x_i-1le x_i-x_i-1$? if $x_i=x_i-1$ then?
    – communnites
    Sep 1 at 8:57











  • and My question right is $sum_i=1^n^2dfrac1i^2-frac12$,not $sum_i=1^n^2dfrac1i^2$
    – communnites
    Sep 1 at 9:02











  • @communnites Because for all natural $m$ we have $sqrtmleq m.$ Also, for $x_nleq n^2$ I proved that $LHSleqsumlimits_i=2^n^2frac1i<sumlimits_i=1^n^2frac1i-frac12.$
    – Michael Rozenberg
    Sep 1 at 9:39













up vote
2
down vote










up vote
2
down vote









I think the following can help.



Let for $n>1$ we have $x_nleq n^2$.



Thus, $$sum_i=1^nfracsqrtx_i-x_i-1x_ileqsum_i=1^nfracx_i-x_i-1x_i=sum_i=1^nsum_j=1^x_i-x_i-1frac1x_ileq$$
$$leqsum_i=1^nsum_j=1^x_i-x_i-1frac1x_i-1+j=sum_i=x_0+1^x_nfrac1ileqsum_i=2^n^2frac1i<sum_i=1^n^2frac1i-frac12.$$
If first $x_k>n^2$.



Thus, $$sum_i=k^nfracsqrtx_i-x_i-1x_i<sum_i=k^nfrac1sqrtx_i<sum_i=k^nfrac1n<1$$ and
$$sum_i=1^k-1fracsqrtx_i-x_i-1x_i<sum_i=2^n^2frac1i.$$






share|cite|improve this answer














I think the following can help.



Let for $n>1$ we have $x_nleq n^2$.



Thus, $$sum_i=1^nfracsqrtx_i-x_i-1x_ileqsum_i=1^nfracx_i-x_i-1x_i=sum_i=1^nsum_j=1^x_i-x_i-1frac1x_ileq$$
$$leqsum_i=1^nsum_j=1^x_i-x_i-1frac1x_i-1+j=sum_i=x_0+1^x_nfrac1ileqsum_i=2^n^2frac1i<sum_i=1^n^2frac1i-frac12.$$
If first $x_k>n^2$.



Thus, $$sum_i=k^nfracsqrtx_i-x_i-1x_i<sum_i=k^nfrac1sqrtx_i<sum_i=k^nfrac1n<1$$ and
$$sum_i=1^k-1fracsqrtx_i-x_i-1x_i<sum_i=2^n^2frac1i.$$







share|cite|improve this answer














share|cite|improve this answer



share|cite|improve this answer








edited Sep 1 at 20:25

























answered Sep 1 at 6:21









Michael Rozenberg

89.1k1582179




89.1k1582179











  • then How to do it?
    – communnites
    Sep 1 at 8:35










  • @communnites For $x_nleq n^2$ it's proven. Try to end the proof if $x_k>n^2$ for some $k<n$, where $x_k-1leq n^2.$ The biggest part of the work I made already.
    – Michael Rozenberg
    Sep 1 at 8:51











  • why $sqrtx_i-x_i-1le x_i-x_i-1$? if $x_i=x_i-1$ then?
    – communnites
    Sep 1 at 8:57











  • and My question right is $sum_i=1^n^2dfrac1i^2-frac12$,not $sum_i=1^n^2dfrac1i^2$
    – communnites
    Sep 1 at 9:02











  • @communnites Because for all natural $m$ we have $sqrtmleq m.$ Also, for $x_nleq n^2$ I proved that $LHSleqsumlimits_i=2^n^2frac1i<sumlimits_i=1^n^2frac1i-frac12.$
    – Michael Rozenberg
    Sep 1 at 9:39

















  • then How to do it?
    – communnites
    Sep 1 at 8:35










  • @communnites For $x_nleq n^2$ it's proven. Try to end the proof if $x_k>n^2$ for some $k<n$, where $x_k-1leq n^2.$ The biggest part of the work I made already.
    – Michael Rozenberg
    Sep 1 at 8:51











  • why $sqrtx_i-x_i-1le x_i-x_i-1$? if $x_i=x_i-1$ then?
    – communnites
    Sep 1 at 8:57











  • and My question right is $sum_i=1^n^2dfrac1i^2-frac12$,not $sum_i=1^n^2dfrac1i^2$
    – communnites
    Sep 1 at 9:02











  • @communnites Because for all natural $m$ we have $sqrtmleq m.$ Also, for $x_nleq n^2$ I proved that $LHSleqsumlimits_i=2^n^2frac1i<sumlimits_i=1^n^2frac1i-frac12.$
    – Michael Rozenberg
    Sep 1 at 9:39
















then How to do it?
– communnites
Sep 1 at 8:35




then How to do it?
– communnites
Sep 1 at 8:35












@communnites For $x_nleq n^2$ it's proven. Try to end the proof if $x_k>n^2$ for some $k<n$, where $x_k-1leq n^2.$ The biggest part of the work I made already.
– Michael Rozenberg
Sep 1 at 8:51





@communnites For $x_nleq n^2$ it's proven. Try to end the proof if $x_k>n^2$ for some $k<n$, where $x_k-1leq n^2.$ The biggest part of the work I made already.
– Michael Rozenberg
Sep 1 at 8:51













why $sqrtx_i-x_i-1le x_i-x_i-1$? if $x_i=x_i-1$ then?
– communnites
Sep 1 at 8:57





why $sqrtx_i-x_i-1le x_i-x_i-1$? if $x_i=x_i-1$ then?
– communnites
Sep 1 at 8:57













and My question right is $sum_i=1^n^2dfrac1i^2-frac12$,not $sum_i=1^n^2dfrac1i^2$
– communnites
Sep 1 at 9:02





and My question right is $sum_i=1^n^2dfrac1i^2-frac12$,not $sum_i=1^n^2dfrac1i^2$
– communnites
Sep 1 at 9:02













@communnites Because for all natural $m$ we have $sqrtmleq m.$ Also, for $x_nleq n^2$ I proved that $LHSleqsumlimits_i=2^n^2frac1i<sumlimits_i=1^n^2frac1i-frac12.$
– Michael Rozenberg
Sep 1 at 9:39





@communnites Because for all natural $m$ we have $sqrtmleq m.$ Also, for $x_nleq n^2$ I proved that $LHSleqsumlimits_i=2^n^2frac1i<sumlimits_i=1^n^2frac1i-frac12.$
– Michael Rozenberg
Sep 1 at 9:39











up vote
-1
down vote













I am going to treat $x_i$ as a real variable. Then, we can differentiate the terms with respect to $x_i$:



$$fracddx_ifracsqrtx_i-x_i-1x_i$$



It is easy to show that the derivative is $ 0$ when $x_i=2x_i-1$, and it is also easy to show that it is a maximum (take the 2nd derivave, or just look at the cases $x_i=x_i-1$ and $x_irightarrowinfty $ .



So replacing $x_i=2x_i-1$ we get that each term becomes $frac12sqrtx_i-1$.



Note too, that when using this values for $x_i$, the smaller the values we choose, the term becomes bigger, so there is no doubt that the sum is maxed when we use $x_i=2^i$, therefore:



$$sum_i=1^n fracsqrtx_i-x_i-1x_i leqsum_i=0^n frac12sqrt2^i $$



So if show the worst case we are done. This last series can be easily solved as a geometric one, so the problem is now reduced to showing that:



$$(1+frac 1sqrt2)(1-(frac 1sqrt2)^n+1) leq sum_i=1^n^2 frac 1i - frac12 quad mathbf(1) $$



I will do it by induction. The case $n=1$ is easy. Now we take $(1)$ as our hypothesis, and we want to prove that if the case for $n$ is true, then the case for $n+1$ is true. Starting by the $n+1$ case in the right hand side, we have that:



$$sum_i=1^(n+1)^2 frac 1i - frac12 = sum_i=1^n^2 frac 1i - frac12 +sum_i=n^2+1^(n+1)^2 frac 1i $$



Now we use $(1)$ with the sum up to $n^2$ (and the negative one half), and since we can replace any quantity with a smaller one and the inequality will still be valid, I will replace the latter sum by the number of terms of the sum (expand $(n+1)^2$ to see that the number of terms is $2n+1$ ) multiplied by the smallest term, so now we have to show that:



$$ (1+frac 1sqrt2)(1-(frac 1sqrt2)^n+1)+ frac2n+1(n+1)^2geq (1+frac 1sqrt2)(1-(frac 1sqrt2)^n+2) $$



Rearranging:



$$frac2n+1(n+1)^2geq frac 12 (frac1sqrt2)^n+1 quad mathbf(2)$$



And here I will use a second induction. First, you show that the inequality holds for $n in text1,2,3,4,5$, and using $(2)$ as the hypothesis, we work the left hand side for the case $n+1$ as follows:



$$frac2n+3(n+2)^2 gt frac2n+3sqrt2(n+1)^2 $$



This last inequality, holds for $n gt 5$, because for these values of n, we have that $sqrt[4]2(n+1) gt n+2 $ (That's why I trated the other 5 cases separately). So now we can replace the hypothesis to get:



$$frac2n+3sqrt2(n+1)^2 = frac 12sqrt2 (frac 1sqrt2)^n+1 + frac22sqrt2(n+1)^2 gt frac 12 (frac 1sqrt2)^n+2 $$



Therefore,



$$frac2n+3(n+2)^2 gt frac 12 (frac 1sqrt2)^n+2 $$



This means the $n+1$ case is true, then the hypothesis $(2)$ is true for all natural $n$, and the same goes to the hypothesis $(1)$ and the proof is completed:



$$sum_i=1^n fracsqrtx_i-x_i-1x_i leqsum_i=0^n frac12sqrt2^i leq sum_i=1^n^2 frac 1i - frac12 $$






share|cite|improve this answer




















  • it's right? Thanks
    – communnites
    Sep 2 at 2:01














up vote
-1
down vote













I am going to treat $x_i$ as a real variable. Then, we can differentiate the terms with respect to $x_i$:



$$fracddx_ifracsqrtx_i-x_i-1x_i$$



It is easy to show that the derivative is $ 0$ when $x_i=2x_i-1$, and it is also easy to show that it is a maximum (take the 2nd derivave, or just look at the cases $x_i=x_i-1$ and $x_irightarrowinfty $ .



So replacing $x_i=2x_i-1$ we get that each term becomes $frac12sqrtx_i-1$.



Note too, that when using this values for $x_i$, the smaller the values we choose, the term becomes bigger, so there is no doubt that the sum is maxed when we use $x_i=2^i$, therefore:



$$sum_i=1^n fracsqrtx_i-x_i-1x_i leqsum_i=0^n frac12sqrt2^i $$



So if show the worst case we are done. This last series can be easily solved as a geometric one, so the problem is now reduced to showing that:



$$(1+frac 1sqrt2)(1-(frac 1sqrt2)^n+1) leq sum_i=1^n^2 frac 1i - frac12 quad mathbf(1) $$



I will do it by induction. The case $n=1$ is easy. Now we take $(1)$ as our hypothesis, and we want to prove that if the case for $n$ is true, then the case for $n+1$ is true. Starting by the $n+1$ case in the right hand side, we have that:



$$sum_i=1^(n+1)^2 frac 1i - frac12 = sum_i=1^n^2 frac 1i - frac12 +sum_i=n^2+1^(n+1)^2 frac 1i $$



Now we use $(1)$ with the sum up to $n^2$ (and the negative one half), and since we can replace any quantity with a smaller one and the inequality will still be valid, I will replace the latter sum by the number of terms of the sum (expand $(n+1)^2$ to see that the number of terms is $2n+1$ ) multiplied by the smallest term, so now we have to show that:



$$ (1+frac 1sqrt2)(1-(frac 1sqrt2)^n+1)+ frac2n+1(n+1)^2geq (1+frac 1sqrt2)(1-(frac 1sqrt2)^n+2) $$



Rearranging:



$$frac2n+1(n+1)^2geq frac 12 (frac1sqrt2)^n+1 quad mathbf(2)$$



And here I will use a second induction. First, you show that the inequality holds for $n in text1,2,3,4,5$, and using $(2)$ as the hypothesis, we work the left hand side for the case $n+1$ as follows:



$$frac2n+3(n+2)^2 gt frac2n+3sqrt2(n+1)^2 $$



This last inequality, holds for $n gt 5$, because for these values of n, we have that $sqrt[4]2(n+1) gt n+2 $ (That's why I trated the other 5 cases separately). So now we can replace the hypothesis to get:



$$frac2n+3sqrt2(n+1)^2 = frac 12sqrt2 (frac 1sqrt2)^n+1 + frac22sqrt2(n+1)^2 gt frac 12 (frac 1sqrt2)^n+2 $$



Therefore,



$$frac2n+3(n+2)^2 gt frac 12 (frac 1sqrt2)^n+2 $$



This means the $n+1$ case is true, then the hypothesis $(2)$ is true for all natural $n$, and the same goes to the hypothesis $(1)$ and the proof is completed:



$$sum_i=1^n fracsqrtx_i-x_i-1x_i leqsum_i=0^n frac12sqrt2^i leq sum_i=1^n^2 frac 1i - frac12 $$






share|cite|improve this answer




















  • it's right? Thanks
    – communnites
    Sep 2 at 2:01












up vote
-1
down vote










up vote
-1
down vote









I am going to treat $x_i$ as a real variable. Then, we can differentiate the terms with respect to $x_i$:



$$fracddx_ifracsqrtx_i-x_i-1x_i$$



It is easy to show that the derivative is $ 0$ when $x_i=2x_i-1$, and it is also easy to show that it is a maximum (take the 2nd derivave, or just look at the cases $x_i=x_i-1$ and $x_irightarrowinfty $ .



So replacing $x_i=2x_i-1$ we get that each term becomes $frac12sqrtx_i-1$.



Note too, that when using this values for $x_i$, the smaller the values we choose, the term becomes bigger, so there is no doubt that the sum is maxed when we use $x_i=2^i$, therefore:



$$sum_i=1^n fracsqrtx_i-x_i-1x_i leqsum_i=0^n frac12sqrt2^i $$



So if show the worst case we are done. This last series can be easily solved as a geometric one, so the problem is now reduced to showing that:



$$(1+frac 1sqrt2)(1-(frac 1sqrt2)^n+1) leq sum_i=1^n^2 frac 1i - frac12 quad mathbf(1) $$



I will do it by induction. The case $n=1$ is easy. Now we take $(1)$ as our hypothesis, and we want to prove that if the case for $n$ is true, then the case for $n+1$ is true. Starting by the $n+1$ case in the right hand side, we have that:



$$sum_i=1^(n+1)^2 frac 1i - frac12 = sum_i=1^n^2 frac 1i - frac12 +sum_i=n^2+1^(n+1)^2 frac 1i $$



Now we use $(1)$ with the sum up to $n^2$ (and the negative one half), and since we can replace any quantity with a smaller one and the inequality will still be valid, I will replace the latter sum by the number of terms of the sum (expand $(n+1)^2$ to see that the number of terms is $2n+1$ ) multiplied by the smallest term, so now we have to show that:



$$ (1+frac 1sqrt2)(1-(frac 1sqrt2)^n+1)+ frac2n+1(n+1)^2geq (1+frac 1sqrt2)(1-(frac 1sqrt2)^n+2) $$



Rearranging:



$$frac2n+1(n+1)^2geq frac 12 (frac1sqrt2)^n+1 quad mathbf(2)$$



And here I will use a second induction. First, you show that the inequality holds for $n in text1,2,3,4,5$, and using $(2)$ as the hypothesis, we work the left hand side for the case $n+1$ as follows:



$$frac2n+3(n+2)^2 gt frac2n+3sqrt2(n+1)^2 $$



This last inequality, holds for $n gt 5$, because for these values of n, we have that $sqrt[4]2(n+1) gt n+2 $ (That's why I trated the other 5 cases separately). So now we can replace the hypothesis to get:



$$frac2n+3sqrt2(n+1)^2 = frac 12sqrt2 (frac 1sqrt2)^n+1 + frac22sqrt2(n+1)^2 gt frac 12 (frac 1sqrt2)^n+2 $$



Therefore,



$$frac2n+3(n+2)^2 gt frac 12 (frac 1sqrt2)^n+2 $$



This means the $n+1$ case is true, then the hypothesis $(2)$ is true for all natural $n$, and the same goes to the hypothesis $(1)$ and the proof is completed:



$$sum_i=1^n fracsqrtx_i-x_i-1x_i leqsum_i=0^n frac12sqrt2^i leq sum_i=1^n^2 frac 1i - frac12 $$






share|cite|improve this answer












I am going to treat $x_i$ as a real variable. Then, we can differentiate the terms with respect to $x_i$:



$$fracddx_ifracsqrtx_i-x_i-1x_i$$



It is easy to show that the derivative is $ 0$ when $x_i=2x_i-1$, and it is also easy to show that it is a maximum (take the 2nd derivave, or just look at the cases $x_i=x_i-1$ and $x_irightarrowinfty $ .



So replacing $x_i=2x_i-1$ we get that each term becomes $frac12sqrtx_i-1$.



Note too, that when using this values for $x_i$, the smaller the values we choose, the term becomes bigger, so there is no doubt that the sum is maxed when we use $x_i=2^i$, therefore:



$$sum_i=1^n fracsqrtx_i-x_i-1x_i leqsum_i=0^n frac12sqrt2^i $$



So if show the worst case we are done. This last series can be easily solved as a geometric one, so the problem is now reduced to showing that:



$$(1+frac 1sqrt2)(1-(frac 1sqrt2)^n+1) leq sum_i=1^n^2 frac 1i - frac12 quad mathbf(1) $$



I will do it by induction. The case $n=1$ is easy. Now we take $(1)$ as our hypothesis, and we want to prove that if the case for $n$ is true, then the case for $n+1$ is true. Starting by the $n+1$ case in the right hand side, we have that:



$$sum_i=1^(n+1)^2 frac 1i - frac12 = sum_i=1^n^2 frac 1i - frac12 +sum_i=n^2+1^(n+1)^2 frac 1i $$



Now we use $(1)$ with the sum up to $n^2$ (and the negative one half), and since we can replace any quantity with a smaller one and the inequality will still be valid, I will replace the latter sum by the number of terms of the sum (expand $(n+1)^2$ to see that the number of terms is $2n+1$ ) multiplied by the smallest term, so now we have to show that:



$$ (1+frac 1sqrt2)(1-(frac 1sqrt2)^n+1)+ frac2n+1(n+1)^2geq (1+frac 1sqrt2)(1-(frac 1sqrt2)^n+2) $$



Rearranging:



$$frac2n+1(n+1)^2geq frac 12 (frac1sqrt2)^n+1 quad mathbf(2)$$



And here I will use a second induction. First, you show that the inequality holds for $n in text1,2,3,4,5$, and using $(2)$ as the hypothesis, we work the left hand side for the case $n+1$ as follows:



$$frac2n+3(n+2)^2 gt frac2n+3sqrt2(n+1)^2 $$



This last inequality, holds for $n gt 5$, because for these values of n, we have that $sqrt[4]2(n+1) gt n+2 $ (That's why I trated the other 5 cases separately). So now we can replace the hypothesis to get:



$$frac2n+3sqrt2(n+1)^2 = frac 12sqrt2 (frac 1sqrt2)^n+1 + frac22sqrt2(n+1)^2 gt frac 12 (frac 1sqrt2)^n+2 $$



Therefore,



$$frac2n+3(n+2)^2 gt frac 12 (frac 1sqrt2)^n+2 $$



This means the $n+1$ case is true, then the hypothesis $(2)$ is true for all natural $n$, and the same goes to the hypothesis $(1)$ and the proof is completed:



$$sum_i=1^n fracsqrtx_i-x_i-1x_i leqsum_i=0^n frac12sqrt2^i leq sum_i=1^n^2 frac 1i - frac12 $$







share|cite|improve this answer












share|cite|improve this answer



share|cite|improve this answer










answered Sep 1 at 23:01









Francisco Abusleme

1624




1624











  • it's right? Thanks
    – communnites
    Sep 2 at 2:01
















  • it's right? Thanks
    – communnites
    Sep 2 at 2:01















it's right? Thanks
– communnites
Sep 2 at 2:01




it's right? Thanks
– communnites
Sep 2 at 2:01

















 

draft saved


draft discarded















































 


draft saved


draft discarded














StackExchange.ready(
function ()
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fmath.stackexchange.com%2fquestions%2f2901361%2finquality-sum-i-1n-frac-sqrtx-i-x-i-1x-i-le-sum-i-1n2-f%23new-answer', 'question_page');

);

Post as a guest













































































這個網誌中的熱門文章

How to combine Bézier curves to a surface?

Mutual Information Always Non-negative

Why am i infinitely getting the same tweet with the Twitter Search API?