An interpolation inequality for Fourier cosine series

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











up vote
1
down vote

favorite
3












This question arises as a follow-up to MSE2874264.
Let us assume to have $fin L^2(-pi,pi)$ defined by
$$ f(x) = sum_ngeq 1 c_n cos(nx) $$
where the coefficients $c_n$ are non-negative, decreasing to zero and such that $c_n=Oleft(frac1nright)$.




Is it true that for any $ain(1,+infty)$ the quantity
$ c_a stackreltextdef=frac1piint_-pi^pif(x)cos(ax),dx $ belongs to the interval $left[c_lfloor arfloor+1,c_lfloor a rfloorright]$? In such a case, what conditions on $f$ ensure that $c_a$ is a decreasing function?




We may easily check that
$$ c_a = sum_ngeq 1frac2a(-1)^n sin(pi a)pi(a^2-n^2)c_n $$
so if $ainmathbbN^+$ we have that $c_a$ agrees with the coefficient $c_lfloor arfloor$ from the Fourier cosine series of $f$.
If that is not the case, an efficient strategy for bounding the previous series seems to invoke the Poisson summation formula and the fact that the Fourier transform of the $operatornamesinc$ is well-known. Besides that I have not been able to produce sharp inequalities for $c_a$ in terms of $c_lfloor arfloor$ and $c_lfloor arfloor+1$ only.







share|cite|improve this question


















  • 2




    It might be relevant to point out that by the Paley-Wiener theorem the Fourier transform of a compact-supported function is an entire function with order $1$, so the question is probably equivalent to: if we know that an entire function is decreasing to zero over $mathbbN^+$, what can be said about its behaviour over the half-lines contained in $mathbbR^+$?
    – Jack D'Aurizio♦
    Aug 9 at 19:56










  • Why is the $(-1)^n$ in your formula? When I take your formula, it's the same as $c_a = 2asum_n=1^inftytfrac(-1)^na+noperatornamesinc(a-n)c_n$. Hence, $c_m = 2mtfrac(-1)^m2mc_m = (-1)^mc_m$...
    – amsmath
    Aug 9 at 21:25










  • @amsmath: you forgot the term $frac1acolorred-n$ and $$frac1piint_-pi^picos(ax)cos(nx),dx = frac2a,overbracecolorredcos(npi)^(-1)^nsin(pi a)pi(a^2-n^2).$$
    – Jack D'Aurizio♦
    Aug 9 at 21:33







  • 1




    As a counterexample, check out $f(x) = cos(x)$. Then $c_a = tfrac2aa+1operatornamesinc(a-1)$, which oscillates around $0$.
    – amsmath
    Aug 10 at 0:04






  • 1




    For further possible calculations you might also want to work with the formula $c_a = sum_n=-infty^infty c_operatornamesinc(a-n)$. Here and before, I used the definition $operatornamesinc(x) = tfracsin(pi x)pi x$.
    – amsmath
    Aug 10 at 0:08















up vote
1
down vote

favorite
3












This question arises as a follow-up to MSE2874264.
Let us assume to have $fin L^2(-pi,pi)$ defined by
$$ f(x) = sum_ngeq 1 c_n cos(nx) $$
where the coefficients $c_n$ are non-negative, decreasing to zero and such that $c_n=Oleft(frac1nright)$.




Is it true that for any $ain(1,+infty)$ the quantity
$ c_a stackreltextdef=frac1piint_-pi^pif(x)cos(ax),dx $ belongs to the interval $left[c_lfloor arfloor+1,c_lfloor a rfloorright]$? In such a case, what conditions on $f$ ensure that $c_a$ is a decreasing function?




We may easily check that
$$ c_a = sum_ngeq 1frac2a(-1)^n sin(pi a)pi(a^2-n^2)c_n $$
so if $ainmathbbN^+$ we have that $c_a$ agrees with the coefficient $c_lfloor arfloor$ from the Fourier cosine series of $f$.
If that is not the case, an efficient strategy for bounding the previous series seems to invoke the Poisson summation formula and the fact that the Fourier transform of the $operatornamesinc$ is well-known. Besides that I have not been able to produce sharp inequalities for $c_a$ in terms of $c_lfloor arfloor$ and $c_lfloor arfloor+1$ only.







share|cite|improve this question


















  • 2




    It might be relevant to point out that by the Paley-Wiener theorem the Fourier transform of a compact-supported function is an entire function with order $1$, so the question is probably equivalent to: if we know that an entire function is decreasing to zero over $mathbbN^+$, what can be said about its behaviour over the half-lines contained in $mathbbR^+$?
    – Jack D'Aurizio♦
    Aug 9 at 19:56










  • Why is the $(-1)^n$ in your formula? When I take your formula, it's the same as $c_a = 2asum_n=1^inftytfrac(-1)^na+noperatornamesinc(a-n)c_n$. Hence, $c_m = 2mtfrac(-1)^m2mc_m = (-1)^mc_m$...
    – amsmath
    Aug 9 at 21:25










  • @amsmath: you forgot the term $frac1acolorred-n$ and $$frac1piint_-pi^picos(ax)cos(nx),dx = frac2a,overbracecolorredcos(npi)^(-1)^nsin(pi a)pi(a^2-n^2).$$
    – Jack D'Aurizio♦
    Aug 9 at 21:33







  • 1




    As a counterexample, check out $f(x) = cos(x)$. Then $c_a = tfrac2aa+1operatornamesinc(a-1)$, which oscillates around $0$.
    – amsmath
    Aug 10 at 0:04






  • 1




    For further possible calculations you might also want to work with the formula $c_a = sum_n=-infty^infty c_operatornamesinc(a-n)$. Here and before, I used the definition $operatornamesinc(x) = tfracsin(pi x)pi x$.
    – amsmath
    Aug 10 at 0:08













up vote
1
down vote

favorite
3









up vote
1
down vote

favorite
3






3





This question arises as a follow-up to MSE2874264.
Let us assume to have $fin L^2(-pi,pi)$ defined by
$$ f(x) = sum_ngeq 1 c_n cos(nx) $$
where the coefficients $c_n$ are non-negative, decreasing to zero and such that $c_n=Oleft(frac1nright)$.




Is it true that for any $ain(1,+infty)$ the quantity
$ c_a stackreltextdef=frac1piint_-pi^pif(x)cos(ax),dx $ belongs to the interval $left[c_lfloor arfloor+1,c_lfloor a rfloorright]$? In such a case, what conditions on $f$ ensure that $c_a$ is a decreasing function?




We may easily check that
$$ c_a = sum_ngeq 1frac2a(-1)^n sin(pi a)pi(a^2-n^2)c_n $$
so if $ainmathbbN^+$ we have that $c_a$ agrees with the coefficient $c_lfloor arfloor$ from the Fourier cosine series of $f$.
If that is not the case, an efficient strategy for bounding the previous series seems to invoke the Poisson summation formula and the fact that the Fourier transform of the $operatornamesinc$ is well-known. Besides that I have not been able to produce sharp inequalities for $c_a$ in terms of $c_lfloor arfloor$ and $c_lfloor arfloor+1$ only.







share|cite|improve this question














This question arises as a follow-up to MSE2874264.
Let us assume to have $fin L^2(-pi,pi)$ defined by
$$ f(x) = sum_ngeq 1 c_n cos(nx) $$
where the coefficients $c_n$ are non-negative, decreasing to zero and such that $c_n=Oleft(frac1nright)$.




Is it true that for any $ain(1,+infty)$ the quantity
$ c_a stackreltextdef=frac1piint_-pi^pif(x)cos(ax),dx $ belongs to the interval $left[c_lfloor arfloor+1,c_lfloor a rfloorright]$? In such a case, what conditions on $f$ ensure that $c_a$ is a decreasing function?




We may easily check that
$$ c_a = sum_ngeq 1frac2a(-1)^n sin(pi a)pi(a^2-n^2)c_n $$
so if $ainmathbbN^+$ we have that $c_a$ agrees with the coefficient $c_lfloor arfloor$ from the Fourier cosine series of $f$.
If that is not the case, an efficient strategy for bounding the previous series seems to invoke the Poisson summation formula and the fact that the Fourier transform of the $operatornamesinc$ is well-known. Besides that I have not been able to produce sharp inequalities for $c_a$ in terms of $c_lfloor arfloor$ and $c_lfloor arfloor+1$ only.









share|cite|improve this question













share|cite|improve this question




share|cite|improve this question








edited Aug 9 at 19:57

























asked Aug 9 at 19:44









Jack D'Aurizio♦

271k31266631




271k31266631







  • 2




    It might be relevant to point out that by the Paley-Wiener theorem the Fourier transform of a compact-supported function is an entire function with order $1$, so the question is probably equivalent to: if we know that an entire function is decreasing to zero over $mathbbN^+$, what can be said about its behaviour over the half-lines contained in $mathbbR^+$?
    – Jack D'Aurizio♦
    Aug 9 at 19:56










  • Why is the $(-1)^n$ in your formula? When I take your formula, it's the same as $c_a = 2asum_n=1^inftytfrac(-1)^na+noperatornamesinc(a-n)c_n$. Hence, $c_m = 2mtfrac(-1)^m2mc_m = (-1)^mc_m$...
    – amsmath
    Aug 9 at 21:25










  • @amsmath: you forgot the term $frac1acolorred-n$ and $$frac1piint_-pi^picos(ax)cos(nx),dx = frac2a,overbracecolorredcos(npi)^(-1)^nsin(pi a)pi(a^2-n^2).$$
    – Jack D'Aurizio♦
    Aug 9 at 21:33







  • 1




    As a counterexample, check out $f(x) = cos(x)$. Then $c_a = tfrac2aa+1operatornamesinc(a-1)$, which oscillates around $0$.
    – amsmath
    Aug 10 at 0:04






  • 1




    For further possible calculations you might also want to work with the formula $c_a = sum_n=-infty^infty c_operatornamesinc(a-n)$. Here and before, I used the definition $operatornamesinc(x) = tfracsin(pi x)pi x$.
    – amsmath
    Aug 10 at 0:08













  • 2




    It might be relevant to point out that by the Paley-Wiener theorem the Fourier transform of a compact-supported function is an entire function with order $1$, so the question is probably equivalent to: if we know that an entire function is decreasing to zero over $mathbbN^+$, what can be said about its behaviour over the half-lines contained in $mathbbR^+$?
    – Jack D'Aurizio♦
    Aug 9 at 19:56










  • Why is the $(-1)^n$ in your formula? When I take your formula, it's the same as $c_a = 2asum_n=1^inftytfrac(-1)^na+noperatornamesinc(a-n)c_n$. Hence, $c_m = 2mtfrac(-1)^m2mc_m = (-1)^mc_m$...
    – amsmath
    Aug 9 at 21:25










  • @amsmath: you forgot the term $frac1acolorred-n$ and $$frac1piint_-pi^picos(ax)cos(nx),dx = frac2a,overbracecolorredcos(npi)^(-1)^nsin(pi a)pi(a^2-n^2).$$
    – Jack D'Aurizio♦
    Aug 9 at 21:33







  • 1




    As a counterexample, check out $f(x) = cos(x)$. Then $c_a = tfrac2aa+1operatornamesinc(a-1)$, which oscillates around $0$.
    – amsmath
    Aug 10 at 0:04






  • 1




    For further possible calculations you might also want to work with the formula $c_a = sum_n=-infty^infty c_operatornamesinc(a-n)$. Here and before, I used the definition $operatornamesinc(x) = tfracsin(pi x)pi x$.
    – amsmath
    Aug 10 at 0:08








2




2




It might be relevant to point out that by the Paley-Wiener theorem the Fourier transform of a compact-supported function is an entire function with order $1$, so the question is probably equivalent to: if we know that an entire function is decreasing to zero over $mathbbN^+$, what can be said about its behaviour over the half-lines contained in $mathbbR^+$?
– Jack D'Aurizio♦
Aug 9 at 19:56




It might be relevant to point out that by the Paley-Wiener theorem the Fourier transform of a compact-supported function is an entire function with order $1$, so the question is probably equivalent to: if we know that an entire function is decreasing to zero over $mathbbN^+$, what can be said about its behaviour over the half-lines contained in $mathbbR^+$?
– Jack D'Aurizio♦
Aug 9 at 19:56












Why is the $(-1)^n$ in your formula? When I take your formula, it's the same as $c_a = 2asum_n=1^inftytfrac(-1)^na+noperatornamesinc(a-n)c_n$. Hence, $c_m = 2mtfrac(-1)^m2mc_m = (-1)^mc_m$...
– amsmath
Aug 9 at 21:25




Why is the $(-1)^n$ in your formula? When I take your formula, it's the same as $c_a = 2asum_n=1^inftytfrac(-1)^na+noperatornamesinc(a-n)c_n$. Hence, $c_m = 2mtfrac(-1)^m2mc_m = (-1)^mc_m$...
– amsmath
Aug 9 at 21:25












@amsmath: you forgot the term $frac1acolorred-n$ and $$frac1piint_-pi^picos(ax)cos(nx),dx = frac2a,overbracecolorredcos(npi)^(-1)^nsin(pi a)pi(a^2-n^2).$$
– Jack D'Aurizio♦
Aug 9 at 21:33





@amsmath: you forgot the term $frac1acolorred-n$ and $$frac1piint_-pi^picos(ax)cos(nx),dx = frac2a,overbracecolorredcos(npi)^(-1)^nsin(pi a)pi(a^2-n^2).$$
– Jack D'Aurizio♦
Aug 9 at 21:33





1




1




As a counterexample, check out $f(x) = cos(x)$. Then $c_a = tfrac2aa+1operatornamesinc(a-1)$, which oscillates around $0$.
– amsmath
Aug 10 at 0:04




As a counterexample, check out $f(x) = cos(x)$. Then $c_a = tfrac2aa+1operatornamesinc(a-1)$, which oscillates around $0$.
– amsmath
Aug 10 at 0:04




1




1




For further possible calculations you might also want to work with the formula $c_a = sum_n=-infty^infty c_operatornamesinc(a-n)$. Here and before, I used the definition $operatornamesinc(x) = tfracsin(pi x)pi x$.
– amsmath
Aug 10 at 0:08





For further possible calculations you might also want to work with the formula $c_a = sum_n=-infty^infty c_operatornamesinc(a-n)$. Here and before, I used the definition $operatornamesinc(x) = tfracsin(pi x)pi x$.
– amsmath
Aug 10 at 0:08











1 Answer
1






active

oldest

votes

















up vote
0
down vote













The answer relies in Gibbs phenomenon. As pointed out in the comments below the main question it is not granted that if $ain(n,n+1)$ then $c_ain(c_n+1,c_n)$, but due to the following lemma, the given constraints ensure that $c_a$ cannot lie too far from such interval:




For any $ngeq 4$ and any $xin[0,n]$ we have that
$$ f_n(x)stackreltextdef=sum_k=0^ntextsinc(pi(x-k)) in left[frac2pitextSi(2pi),frac2pitextSi(pi)right]subset[0.902,1.179]$$
where $textsinc(x)=left{beginarraycl1&textif x=0\ fracsin xx&textotherwiseendarrayright.$ and $textSi(x)=int_0^xtextsinc(t),dt.$







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%2f2877625%2fan-interpolation-inequality-for-fourier-cosine-series%23new-answer', 'question_page');

    );

    Post as a guest






























    1 Answer
    1






    active

    oldest

    votes








    1 Answer
    1






    active

    oldest

    votes









    active

    oldest

    votes






    active

    oldest

    votes








    up vote
    0
    down vote













    The answer relies in Gibbs phenomenon. As pointed out in the comments below the main question it is not granted that if $ain(n,n+1)$ then $c_ain(c_n+1,c_n)$, but due to the following lemma, the given constraints ensure that $c_a$ cannot lie too far from such interval:




    For any $ngeq 4$ and any $xin[0,n]$ we have that
    $$ f_n(x)stackreltextdef=sum_k=0^ntextsinc(pi(x-k)) in left[frac2pitextSi(2pi),frac2pitextSi(pi)right]subset[0.902,1.179]$$
    where $textsinc(x)=left{beginarraycl1&textif x=0\ fracsin xx&textotherwiseendarrayright.$ and $textSi(x)=int_0^xtextsinc(t),dt.$







    share|cite|improve this answer
























      up vote
      0
      down vote













      The answer relies in Gibbs phenomenon. As pointed out in the comments below the main question it is not granted that if $ain(n,n+1)$ then $c_ain(c_n+1,c_n)$, but due to the following lemma, the given constraints ensure that $c_a$ cannot lie too far from such interval:




      For any $ngeq 4$ and any $xin[0,n]$ we have that
      $$ f_n(x)stackreltextdef=sum_k=0^ntextsinc(pi(x-k)) in left[frac2pitextSi(2pi),frac2pitextSi(pi)right]subset[0.902,1.179]$$
      where $textsinc(x)=left{beginarraycl1&textif x=0\ fracsin xx&textotherwiseendarrayright.$ and $textSi(x)=int_0^xtextsinc(t),dt.$







      share|cite|improve this answer






















        up vote
        0
        down vote










        up vote
        0
        down vote









        The answer relies in Gibbs phenomenon. As pointed out in the comments below the main question it is not granted that if $ain(n,n+1)$ then $c_ain(c_n+1,c_n)$, but due to the following lemma, the given constraints ensure that $c_a$ cannot lie too far from such interval:




        For any $ngeq 4$ and any $xin[0,n]$ we have that
        $$ f_n(x)stackreltextdef=sum_k=0^ntextsinc(pi(x-k)) in left[frac2pitextSi(2pi),frac2pitextSi(pi)right]subset[0.902,1.179]$$
        where $textsinc(x)=left{beginarraycl1&textif x=0\ fracsin xx&textotherwiseendarrayright.$ and $textSi(x)=int_0^xtextsinc(t),dt.$







        share|cite|improve this answer












        The answer relies in Gibbs phenomenon. As pointed out in the comments below the main question it is not granted that if $ain(n,n+1)$ then $c_ain(c_n+1,c_n)$, but due to the following lemma, the given constraints ensure that $c_a$ cannot lie too far from such interval:




        For any $ngeq 4$ and any $xin[0,n]$ we have that
        $$ f_n(x)stackreltextdef=sum_k=0^ntextsinc(pi(x-k)) in left[frac2pitextSi(2pi),frac2pitextSi(pi)right]subset[0.902,1.179]$$
        where $textsinc(x)=left{beginarraycl1&textif x=0\ fracsin xx&textotherwiseendarrayright.$ and $textSi(x)=int_0^xtextsinc(t),dt.$








        share|cite|improve this answer












        share|cite|improve this answer



        share|cite|improve this answer










        answered Aug 13 at 6:39









        Jack D'Aurizio♦

        271k31266631




        271k31266631






















             

            draft saved


            draft discarded


























             


            draft saved


            draft discarded














            StackExchange.ready(
            function ()
            StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fmath.stackexchange.com%2fquestions%2f2877625%2fan-interpolation-inequality-for-fourier-cosine-series%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?

            Carbon dioxide

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