Proving that $int_0^1 frac{arctan x}{x}lnleft(frac{1+x^2}{(1-x)^2}right)dx=frac{pi^3}{16}$
$begingroup$
The following integral was proposed by Cornel Ioan Valean and appeared as Problem $12054$ in the American Mathematical Monthly earlier this year.
Prove $$int_0^1 frac{arctan x}{x}lnleft(frac{1+x^2}{(1-x)^2}right)dx=frac{pi^3}{16}$$
I had small tries for it, such as:
Letting $x=tan t$ which gives:
$$I=int_0^1 frac{arctan x}{x}lnleft(frac{1+x^2}{(1-x)^2}right)dx=-int_0^frac{pi}{4}frac{t}{sin t}ln(1-sin(2t))dt=$$
$$overset{2t=x}=-frac12 underset{=J}{int_0^frac{pi}{2}frac{x}{sin x} ln(1-sin x)dx}=-frac12 int_0^frac{pi}{2} xln(1-sin x) left(lnleft(tan frac{x}{2}right)right)'dx=$$
$$overset{IBP}=frac12int_0^frac{pi}{2} lnleft(1-sin xright)lnleft(tan frac{x}{2}right)dx+frac12 int_0^frac{pi}{2} frac{xcos x}{sin x-1}lnleft(tan frac{x}{2}right)dx$$
Or to employ Feynman's trick for the first integral $(J)$ in the second row. $$J(t)=int_0^frac{pi}{2} frac{xln(1-tsin x)}{sin x}dxRightarrow J'(t)=int_0^frac{pi}{2} frac{x}{1-tsin x}dx$$ But even so I don't see a how to obtain a closed from for the last one. Also with a different parameter: $$J(t)=int_0^frac{pi}{2} frac{text{arccot} (t cot x)ln(1-sin x)}{sin x}dx$$
$$Rightarrow J'(t)=-int_0^frac{pi}{2} frac{ln(1-sin x)cos x}{1+t^2 cot^2x}frac{dx}{sin^2x}overset{sin x=y}=int_0^1 frac{ln(1-y)}{1+t^2left(1-frac{1}{y^2}right)}frac{dy}{y^2}$$
Also from here we have the following relation:
$$int_0^1 frac{arctan x ln(1+x^2)}{x} dx =frac23 int_0^1 frac{arctan x ln(1+x)}{x}dx$$
Thus we can rewrite the integral as:
$$I=frac23 int_0^1 frac{arctan x ln(1+x)}{x}dx -2int_0^1 frac{arctan x ln(1-x)}{x}dx$$
$$=frac23 int_0^1 int_0^1 frac{ln(1+x)-3ln(1-x)}{1+x^2y^2}dydx=frac23 int_0^1 int_t^1 frac{ln(1+x)-3ln(1-x)}{1+t^2}dxdt $$
Another option might be to rewrite:
$$lnleft(frac{1+x^2}{(1-x)^2}right)= lnleft(frac{1+x}{1-x}right)+lnleft(frac{1+x^2}{1-x^2}right)$$
$$Rightarrow I= int_0^1 frac{arctan x}{x}lnleft(frac{1+x}{1-x}right)dx+int_0^1 frac{arctan x}{x}lnleft(frac{1+x^2}{1-x^2}right)dx$$
And now to use the power expansion of the log functions inside to obtain:
$$I=sum_{n=0}^infty frac{2}{2n+1}int_0^1 frac{arctan x}{x} , left(x^{2n+1}+x^{4n+2}right)dx=sum_{n=0}^infty frac{2}{2n+1}int_0^1int_0^1 frac{left(x^{2n+1}+x^{4n+2}right)}{1+y^2x^2}dydx$$
In the meantime I found one nice solution by Roberto Tauraso here.
This seems like an awesome integral and I would like to learn more so I am searching for more approaches.
Would any of you who also already solve it and submitted the answer to the AMM or know how to solve this integral kindly share the solution here?
Edit: Another impressive solution due to Yaghoub Sharifi is found here.
integration definite-integrals logarithms closed-form
$endgroup$
add a comment |
$begingroup$
The following integral was proposed by Cornel Ioan Valean and appeared as Problem $12054$ in the American Mathematical Monthly earlier this year.
Prove $$int_0^1 frac{arctan x}{x}lnleft(frac{1+x^2}{(1-x)^2}right)dx=frac{pi^3}{16}$$
I had small tries for it, such as:
Letting $x=tan t$ which gives:
$$I=int_0^1 frac{arctan x}{x}lnleft(frac{1+x^2}{(1-x)^2}right)dx=-int_0^frac{pi}{4}frac{t}{sin t}ln(1-sin(2t))dt=$$
$$overset{2t=x}=-frac12 underset{=J}{int_0^frac{pi}{2}frac{x}{sin x} ln(1-sin x)dx}=-frac12 int_0^frac{pi}{2} xln(1-sin x) left(lnleft(tan frac{x}{2}right)right)'dx=$$
$$overset{IBP}=frac12int_0^frac{pi}{2} lnleft(1-sin xright)lnleft(tan frac{x}{2}right)dx+frac12 int_0^frac{pi}{2} frac{xcos x}{sin x-1}lnleft(tan frac{x}{2}right)dx$$
Or to employ Feynman's trick for the first integral $(J)$ in the second row. $$J(t)=int_0^frac{pi}{2} frac{xln(1-tsin x)}{sin x}dxRightarrow J'(t)=int_0^frac{pi}{2} frac{x}{1-tsin x}dx$$ But even so I don't see a how to obtain a closed from for the last one. Also with a different parameter: $$J(t)=int_0^frac{pi}{2} frac{text{arccot} (t cot x)ln(1-sin x)}{sin x}dx$$
$$Rightarrow J'(t)=-int_0^frac{pi}{2} frac{ln(1-sin x)cos x}{1+t^2 cot^2x}frac{dx}{sin^2x}overset{sin x=y}=int_0^1 frac{ln(1-y)}{1+t^2left(1-frac{1}{y^2}right)}frac{dy}{y^2}$$
Also from here we have the following relation:
$$int_0^1 frac{arctan x ln(1+x^2)}{x} dx =frac23 int_0^1 frac{arctan x ln(1+x)}{x}dx$$
Thus we can rewrite the integral as:
$$I=frac23 int_0^1 frac{arctan x ln(1+x)}{x}dx -2int_0^1 frac{arctan x ln(1-x)}{x}dx$$
$$=frac23 int_0^1 int_0^1 frac{ln(1+x)-3ln(1-x)}{1+x^2y^2}dydx=frac23 int_0^1 int_t^1 frac{ln(1+x)-3ln(1-x)}{1+t^2}dxdt $$
Another option might be to rewrite:
$$lnleft(frac{1+x^2}{(1-x)^2}right)= lnleft(frac{1+x}{1-x}right)+lnleft(frac{1+x^2}{1-x^2}right)$$
$$Rightarrow I= int_0^1 frac{arctan x}{x}lnleft(frac{1+x}{1-x}right)dx+int_0^1 frac{arctan x}{x}lnleft(frac{1+x^2}{1-x^2}right)dx$$
And now to use the power expansion of the log functions inside to obtain:
$$I=sum_{n=0}^infty frac{2}{2n+1}int_0^1 frac{arctan x}{x} , left(x^{2n+1}+x^{4n+2}right)dx=sum_{n=0}^infty frac{2}{2n+1}int_0^1int_0^1 frac{left(x^{2n+1}+x^{4n+2}right)}{1+y^2x^2}dydx$$
In the meantime I found one nice solution by Roberto Tauraso here.
This seems like an awesome integral and I would like to learn more so I am searching for more approaches.
Would any of you who also already solve it and submitted the answer to the AMM or know how to solve this integral kindly share the solution here?
Edit: Another impressive solution due to Yaghoub Sharifi is found here.
integration definite-integrals logarithms closed-form
$endgroup$
3
$begingroup$
I was able to break it down to an evaluation of harmonic sums $$I=frac{3pi^3}{32}-sum_{n=0}^{infty}frac{frac12left[H_{n/2}-H_{(n-1)/2}right]+frac14left[H_{n+1/4}-H_{n-1/4}right]}{(2n+1)^2}$$ the latter sum should equal $pi^3/32$ which seems to work out numerically but honestly speaking I am lost from hereon. Using the well-known result $beta(3)=pi^3/32$ one could conjecture that the combination of harmonic sums has to come out equal to $(-1)^n/(2n+1)$ in order to complete the representation of $beta(3)$.
$endgroup$
– mrtaurho
Dec 24 '18 at 1:48
1
$begingroup$
I would say this solution here is quite impressive and convincing.
$endgroup$
– mrtaurho
Dec 25 '18 at 15:38
add a comment |
$begingroup$
The following integral was proposed by Cornel Ioan Valean and appeared as Problem $12054$ in the American Mathematical Monthly earlier this year.
Prove $$int_0^1 frac{arctan x}{x}lnleft(frac{1+x^2}{(1-x)^2}right)dx=frac{pi^3}{16}$$
I had small tries for it, such as:
Letting $x=tan t$ which gives:
$$I=int_0^1 frac{arctan x}{x}lnleft(frac{1+x^2}{(1-x)^2}right)dx=-int_0^frac{pi}{4}frac{t}{sin t}ln(1-sin(2t))dt=$$
$$overset{2t=x}=-frac12 underset{=J}{int_0^frac{pi}{2}frac{x}{sin x} ln(1-sin x)dx}=-frac12 int_0^frac{pi}{2} xln(1-sin x) left(lnleft(tan frac{x}{2}right)right)'dx=$$
$$overset{IBP}=frac12int_0^frac{pi}{2} lnleft(1-sin xright)lnleft(tan frac{x}{2}right)dx+frac12 int_0^frac{pi}{2} frac{xcos x}{sin x-1}lnleft(tan frac{x}{2}right)dx$$
Or to employ Feynman's trick for the first integral $(J)$ in the second row. $$J(t)=int_0^frac{pi}{2} frac{xln(1-tsin x)}{sin x}dxRightarrow J'(t)=int_0^frac{pi}{2} frac{x}{1-tsin x}dx$$ But even so I don't see a how to obtain a closed from for the last one. Also with a different parameter: $$J(t)=int_0^frac{pi}{2} frac{text{arccot} (t cot x)ln(1-sin x)}{sin x}dx$$
$$Rightarrow J'(t)=-int_0^frac{pi}{2} frac{ln(1-sin x)cos x}{1+t^2 cot^2x}frac{dx}{sin^2x}overset{sin x=y}=int_0^1 frac{ln(1-y)}{1+t^2left(1-frac{1}{y^2}right)}frac{dy}{y^2}$$
Also from here we have the following relation:
$$int_0^1 frac{arctan x ln(1+x^2)}{x} dx =frac23 int_0^1 frac{arctan x ln(1+x)}{x}dx$$
Thus we can rewrite the integral as:
$$I=frac23 int_0^1 frac{arctan x ln(1+x)}{x}dx -2int_0^1 frac{arctan x ln(1-x)}{x}dx$$
$$=frac23 int_0^1 int_0^1 frac{ln(1+x)-3ln(1-x)}{1+x^2y^2}dydx=frac23 int_0^1 int_t^1 frac{ln(1+x)-3ln(1-x)}{1+t^2}dxdt $$
Another option might be to rewrite:
$$lnleft(frac{1+x^2}{(1-x)^2}right)= lnleft(frac{1+x}{1-x}right)+lnleft(frac{1+x^2}{1-x^2}right)$$
$$Rightarrow I= int_0^1 frac{arctan x}{x}lnleft(frac{1+x}{1-x}right)dx+int_0^1 frac{arctan x}{x}lnleft(frac{1+x^2}{1-x^2}right)dx$$
And now to use the power expansion of the log functions inside to obtain:
$$I=sum_{n=0}^infty frac{2}{2n+1}int_0^1 frac{arctan x}{x} , left(x^{2n+1}+x^{4n+2}right)dx=sum_{n=0}^infty frac{2}{2n+1}int_0^1int_0^1 frac{left(x^{2n+1}+x^{4n+2}right)}{1+y^2x^2}dydx$$
In the meantime I found one nice solution by Roberto Tauraso here.
This seems like an awesome integral and I would like to learn more so I am searching for more approaches.
Would any of you who also already solve it and submitted the answer to the AMM or know how to solve this integral kindly share the solution here?
Edit: Another impressive solution due to Yaghoub Sharifi is found here.
integration definite-integrals logarithms closed-form
$endgroup$
The following integral was proposed by Cornel Ioan Valean and appeared as Problem $12054$ in the American Mathematical Monthly earlier this year.
Prove $$int_0^1 frac{arctan x}{x}lnleft(frac{1+x^2}{(1-x)^2}right)dx=frac{pi^3}{16}$$
I had small tries for it, such as:
Letting $x=tan t$ which gives:
$$I=int_0^1 frac{arctan x}{x}lnleft(frac{1+x^2}{(1-x)^2}right)dx=-int_0^frac{pi}{4}frac{t}{sin t}ln(1-sin(2t))dt=$$
$$overset{2t=x}=-frac12 underset{=J}{int_0^frac{pi}{2}frac{x}{sin x} ln(1-sin x)dx}=-frac12 int_0^frac{pi}{2} xln(1-sin x) left(lnleft(tan frac{x}{2}right)right)'dx=$$
$$overset{IBP}=frac12int_0^frac{pi}{2} lnleft(1-sin xright)lnleft(tan frac{x}{2}right)dx+frac12 int_0^frac{pi}{2} frac{xcos x}{sin x-1}lnleft(tan frac{x}{2}right)dx$$
Or to employ Feynman's trick for the first integral $(J)$ in the second row. $$J(t)=int_0^frac{pi}{2} frac{xln(1-tsin x)}{sin x}dxRightarrow J'(t)=int_0^frac{pi}{2} frac{x}{1-tsin x}dx$$ But even so I don't see a how to obtain a closed from for the last one. Also with a different parameter: $$J(t)=int_0^frac{pi}{2} frac{text{arccot} (t cot x)ln(1-sin x)}{sin x}dx$$
$$Rightarrow J'(t)=-int_0^frac{pi}{2} frac{ln(1-sin x)cos x}{1+t^2 cot^2x}frac{dx}{sin^2x}overset{sin x=y}=int_0^1 frac{ln(1-y)}{1+t^2left(1-frac{1}{y^2}right)}frac{dy}{y^2}$$
Also from here we have the following relation:
$$int_0^1 frac{arctan x ln(1+x^2)}{x} dx =frac23 int_0^1 frac{arctan x ln(1+x)}{x}dx$$
Thus we can rewrite the integral as:
$$I=frac23 int_0^1 frac{arctan x ln(1+x)}{x}dx -2int_0^1 frac{arctan x ln(1-x)}{x}dx$$
$$=frac23 int_0^1 int_0^1 frac{ln(1+x)-3ln(1-x)}{1+x^2y^2}dydx=frac23 int_0^1 int_t^1 frac{ln(1+x)-3ln(1-x)}{1+t^2}dxdt $$
Another option might be to rewrite:
$$lnleft(frac{1+x^2}{(1-x)^2}right)= lnleft(frac{1+x}{1-x}right)+lnleft(frac{1+x^2}{1-x^2}right)$$
$$Rightarrow I= int_0^1 frac{arctan x}{x}lnleft(frac{1+x}{1-x}right)dx+int_0^1 frac{arctan x}{x}lnleft(frac{1+x^2}{1-x^2}right)dx$$
And now to use the power expansion of the log functions inside to obtain:
$$I=sum_{n=0}^infty frac{2}{2n+1}int_0^1 frac{arctan x}{x} , left(x^{2n+1}+x^{4n+2}right)dx=sum_{n=0}^infty frac{2}{2n+1}int_0^1int_0^1 frac{left(x^{2n+1}+x^{4n+2}right)}{1+y^2x^2}dydx$$
In the meantime I found one nice solution by Roberto Tauraso here.
This seems like an awesome integral and I would like to learn more so I am searching for more approaches.
Would any of you who also already solve it and submitted the answer to the AMM or know how to solve this integral kindly share the solution here?
Edit: Another impressive solution due to Yaghoub Sharifi is found here.
integration definite-integrals logarithms closed-form
integration definite-integrals logarithms closed-form
edited Dec 27 '18 at 12:24
Zacky
asked Dec 23 '18 at 20:53
ZackyZacky
6,6851958
6,6851958
3
$begingroup$
I was able to break it down to an evaluation of harmonic sums $$I=frac{3pi^3}{32}-sum_{n=0}^{infty}frac{frac12left[H_{n/2}-H_{(n-1)/2}right]+frac14left[H_{n+1/4}-H_{n-1/4}right]}{(2n+1)^2}$$ the latter sum should equal $pi^3/32$ which seems to work out numerically but honestly speaking I am lost from hereon. Using the well-known result $beta(3)=pi^3/32$ one could conjecture that the combination of harmonic sums has to come out equal to $(-1)^n/(2n+1)$ in order to complete the representation of $beta(3)$.
$endgroup$
– mrtaurho
Dec 24 '18 at 1:48
1
$begingroup$
I would say this solution here is quite impressive and convincing.
$endgroup$
– mrtaurho
Dec 25 '18 at 15:38
add a comment |
3
$begingroup$
I was able to break it down to an evaluation of harmonic sums $$I=frac{3pi^3}{32}-sum_{n=0}^{infty}frac{frac12left[H_{n/2}-H_{(n-1)/2}right]+frac14left[H_{n+1/4}-H_{n-1/4}right]}{(2n+1)^2}$$ the latter sum should equal $pi^3/32$ which seems to work out numerically but honestly speaking I am lost from hereon. Using the well-known result $beta(3)=pi^3/32$ one could conjecture that the combination of harmonic sums has to come out equal to $(-1)^n/(2n+1)$ in order to complete the representation of $beta(3)$.
$endgroup$
– mrtaurho
Dec 24 '18 at 1:48
1
$begingroup$
I would say this solution here is quite impressive and convincing.
$endgroup$
– mrtaurho
Dec 25 '18 at 15:38
3
3
$begingroup$
I was able to break it down to an evaluation of harmonic sums $$I=frac{3pi^3}{32}-sum_{n=0}^{infty}frac{frac12left[H_{n/2}-H_{(n-1)/2}right]+frac14left[H_{n+1/4}-H_{n-1/4}right]}{(2n+1)^2}$$ the latter sum should equal $pi^3/32$ which seems to work out numerically but honestly speaking I am lost from hereon. Using the well-known result $beta(3)=pi^3/32$ one could conjecture that the combination of harmonic sums has to come out equal to $(-1)^n/(2n+1)$ in order to complete the representation of $beta(3)$.
$endgroup$
– mrtaurho
Dec 24 '18 at 1:48
$begingroup$
I was able to break it down to an evaluation of harmonic sums $$I=frac{3pi^3}{32}-sum_{n=0}^{infty}frac{frac12left[H_{n/2}-H_{(n-1)/2}right]+frac14left[H_{n+1/4}-H_{n-1/4}right]}{(2n+1)^2}$$ the latter sum should equal $pi^3/32$ which seems to work out numerically but honestly speaking I am lost from hereon. Using the well-known result $beta(3)=pi^3/32$ one could conjecture that the combination of harmonic sums has to come out equal to $(-1)^n/(2n+1)$ in order to complete the representation of $beta(3)$.
$endgroup$
– mrtaurho
Dec 24 '18 at 1:48
1
1
$begingroup$
I would say this solution here is quite impressive and convincing.
$endgroup$
– mrtaurho
Dec 25 '18 at 15:38
$begingroup$
I would say this solution here is quite impressive and convincing.
$endgroup$
– mrtaurho
Dec 25 '18 at 15:38
add a comment |
2 Answers
2
active
oldest
votes
$begingroup$
Another approach,
Perform integration by parts,
begin{align*}
I&=int_0^1 frac{arctan x}{x}lnleft(frac{1+x^2}{(1-x)^2}right),dx\
&=Big[ln (x) lnleft(frac{1+x^2}{(1-x)^2}right)arctan xBig]_0^1 -int_0^1 frac{ln x}{1+x^2}lnleft(frac{1+x^2}{(1-x)^2}right)dx-int_0^1 frac{2(1+x)ln (x)arctan (x)}{(1-x)(1+x^2)}dx\
&=-int_0^1 frac{ln x}{1+x^2}lnleft(frac{1+x^2}{(1-x)^2}right)dx-2int_0^1 frac{(1+x)ln (x)arctan (x)}{(1-x)(1+x^2)}dx\
end{align*}
For $xin [0;1]$ define the function $R$ by,
begin{align*}
R(x)=int_0^x frac{(1+t)ln t}{(1-t)(1+t^2)}dt=int_0^1 frac{x(1+tx)ln (tx)}{(1-tx)(1+t^2x^2)}dt\
end{align*}
Observe that,
begin{align*}
R(1)=int_0^1 frac{tln t}{1+t}dt+int_0^1 frac{ln t}{1-t}dt
end{align*}
Perform integration by parts,
begin{align*}
I&=-int_0^1 frac{ln x}{1+x^2}lnleft(frac{1+x^2}{(1-x)^2}right)dx-2Big[R(x)arctan xBig]_0^1+2int_0^1frac{R(x)}{1+x^2}dx\
&=-int_0^1 frac{ln x}{1+x^2}lnleft(frac{1+x^2}{(1-x)^2}right)dx-frac{pi}{2}R(1)+2int_0^1 int_0^1 frac{x(1+tx)ln (tx)}{(1-tx)(1+t^2x^2)(1+x^2)}dtdx\
&=-int_0^1 frac{ln x}{1+x^2}lnleft(frac{1+x^2}{(1-x)^2}right)dx-frac{pi}{2}R(1)+int_0^1 ln xleft[frac{1}{1+x^2}lnleft(frac{1+t^2x^2}{(1-tx)^2}right)right]_{t=0}^{t=1} dx+\
&int_0^1 ln tleft[frac{1}{1+t^2}lnleft(frac{1+x^2}{(1-tx)^2}right)+frac{2arctan (tx)}{1-t^2}-frac{2tarctan x}{1+t^2}-frac{2tarctan x}{1-t^2}right]_{x=0}^{x=1} dt\
&=-frac{pi }{2}R(1)+ln 2int_0^1 frac{ln t}{1+t^2}dt-2int_0^1 frac{ln (1-t)ln t}{1+t^2}dt+2int_0^1 frac{ln tarctan t}{1-t^2}dt-\
&frac{pi}{2} int_0^1 frac{tln t}{1+t^2}dt-frac{pi}{2} int_0^1frac{tln t}{1-t^2} dt\
end{align*}
For $xin [0;1]$ define the function $S$ by,
begin{align*}
S(x)=int_0^x frac{ln t}{1-t^2}dt=int_0^1 frac{xln(tx)}{1-t^2x^2} dt
end{align*}
Perform integration by parts,
begin{align*}
int_0^1 frac{ln xarctan x}{1-x^2}dx&=Big[S(x)arctan xBig]_0^1-int_0^1 frac{S(x)}{1+x^2}dx\
&=frac{pi}{4}S(1)-int_0^1 int_0^1 frac{xln(tx)} {(1-t^2x^2)(1+x^2)} dtdx\
&=frac{pi}{4}S(1)-frac{1}{2}int_0^1 left[ frac{ln x}{1+x^2}lnleft(frac{1+tx}{1-tx} right)right]_{t=0}^{t=1} dx-\
&frac{1}{2}int_0^1 left[ frac{ln t}{1+t^2}lnleft(frac{1+x^2}{1-t^2x^2} right)right]_{x=0}^{x=1}dt\
&=frac{pi}{4}S(1)-frac{ln 2}{2}int_0^1 frac{ln t}{1+t^2}dt+int_0^1 frac{ln(1-x)ln x}{1+x^2}dx
end{align*}
Therefore,
begin{align*}I&=piint_0^1frac{2tln t}{t^4-1} dtend{align*}
Perform the change of variable $y=t^2$,
begin{align*}I&=frac{1}{2}pi int_0^1 frac{ln y}{y^2-1}dy\
&=frac{1}{2}pitimes frac{3}{4}zeta(2)\
&=frac{pi^3}{16}
end{align*}
$endgroup$
1
$begingroup$
That's impressive, thank you! I've seen you use this approach alot and it's quite useful, let me a few time to understand it's working better.
$endgroup$
– Zacky
Dec 25 '18 at 17:59
$begingroup$
Well done. (+1)
$endgroup$
– Mark Viola
Dec 26 '18 at 4:26
$begingroup$
Very nice solution and $to +1$
$endgroup$
– Claude Leibovici
Dec 26 '18 at 6:07
1
$begingroup$
I compute $int_0^1 F(t,x)ln t,dx$ and $int_0^1 F(t,x)ln x,dt$ and one can compute an antiderivative $U(t,x)$ of $F(t,x)$ wrt $x$, and on the other hand an antiderivative $V(t,x)$ of $F(t,x)$ wrt $t$.
$endgroup$
– FDP
Dec 26 '18 at 17:07
1
$begingroup$
@Zacky: remember in the double integrals you can choose to integrate wrt $x$ or wrt $t$. If there is a factor $ln x$ you don't want to integrate wrt $x$ first. If there is a factor $ln t$ you don't want to integrate wrt $t$ first. And, $ln(tx)=ln x +ln t$
$endgroup$
– FDP
Dec 26 '18 at 17:25
|
show 1 more comment
$begingroup$
Put
begin{equation*}
I=int_{0}^1dfrac{arctan x}{x}lnleft(dfrac{1+x^2}{(1-x)^2}right), mathrm{d}x.
end{equation*}
Via the substitution $ x=dfrac{z}{z+1}$ we get
begin{equation*}
I = int_{0}^{infty}dfrac{arctan frac{z}{z+1}ln(2z^2+2z+1)}{z^2+z}, mathrm{d}z.
end{equation*}
Put
begin{equation*}
log z=ln|z|+iarg z, quad -pi<arg z <pi.
end{equation*}
Then
begin{equation*}
arctan frac{z}{z+1}ln(2z^2+2z+1) = text{Im}left(log^2(1+z+iz)right).
end{equation*}
Consequently
begin{equation*}
I = text{Im}left(int_{0}^{infty}dfrac{log^2(1+z+iz)}{z^2+z}right)mathrm{d}z.
end{equation*}
However, $ log(z) $ is an analytic function in $ text{Re} z>0 $. According to Cauchys integral theorem we get the same value if we integrate along the curve with the parametrization $ z=(1-i)s, s>0 $.
begin{gather*}
I = text{Im}left(int_{0}^{infty}dfrac{ln^2(2s+1)}{s(s+1-is)}, mathrm{d}sright) = int_{0}^{infty}dfrac{ln^2(2s+1)}{2s^2+2s+1}, mathrm{d}s = \[2ex] int_{0}^{infty}dfrac{2ln^2(2s+1)}{(2s+1)^2+1}, mathrm{d}s = [t=2s+1] = \[2ex] int_{1}^{infty}dfrac{ln^2(t)}{t^2+1}, mathrm{d}t =[u= 1/t] = int_{0}^{1}dfrac{ln^2(u)}{u^2+1}, mathrm{d}u.
end{gather*}
Thus
begin{equation*}
2I = int_{0}^{infty}dfrac{ln^2(u)}{u^2+1}, mathrm{d}u
end{equation*}
In order to evaluate this integral we integrate $ dfrac{log^3(z)}{z^2+1} $ along a keyhole contour and use residue calculus.
In this case $ log z =ln |z|+iarg z, quad 0<arg z < 2pi $.
We get
begin{equation*}
I = dfrac{pi^3}{16}.
end{equation*}
$endgroup$
add a comment |
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',
autoActivateHeartbeat: false,
convertImagesToLinks: true,
noModals: true,
showLowRepImageUploadWarning: true,
reputationToPostImages: 10,
bindNavPrevention: true,
postfix: "",
imageUploader: {
brandingHtml: "Powered by u003ca class="icon-imgur-white" href="https://imgur.com/"u003eu003c/au003e",
contentPolicyHtml: "User contributions licensed under u003ca href="https://creativecommons.org/licenses/by-sa/3.0/"u003ecc by-sa 3.0 with attribution requiredu003c/au003e u003ca href="https://stackoverflow.com/legal/content-policy"u003e(content policy)u003c/au003e",
allowUrls: true
},
noCode: true, onDemand: true,
discardSelector: ".discard-answer"
,immediatelyShowMarkdownHelp:true
});
}
});
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
StackExchange.ready(
function () {
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fmath.stackexchange.com%2fquestions%2f3050696%2fproving-that-int-01-frac-arctan-xx-ln-left-frac1x21-x2-rightd%23new-answer', 'question_page');
}
);
Post as a guest
Required, but never shown
2 Answers
2
active
oldest
votes
2 Answers
2
active
oldest
votes
active
oldest
votes
active
oldest
votes
$begingroup$
Another approach,
Perform integration by parts,
begin{align*}
I&=int_0^1 frac{arctan x}{x}lnleft(frac{1+x^2}{(1-x)^2}right),dx\
&=Big[ln (x) lnleft(frac{1+x^2}{(1-x)^2}right)arctan xBig]_0^1 -int_0^1 frac{ln x}{1+x^2}lnleft(frac{1+x^2}{(1-x)^2}right)dx-int_0^1 frac{2(1+x)ln (x)arctan (x)}{(1-x)(1+x^2)}dx\
&=-int_0^1 frac{ln x}{1+x^2}lnleft(frac{1+x^2}{(1-x)^2}right)dx-2int_0^1 frac{(1+x)ln (x)arctan (x)}{(1-x)(1+x^2)}dx\
end{align*}
For $xin [0;1]$ define the function $R$ by,
begin{align*}
R(x)=int_0^x frac{(1+t)ln t}{(1-t)(1+t^2)}dt=int_0^1 frac{x(1+tx)ln (tx)}{(1-tx)(1+t^2x^2)}dt\
end{align*}
Observe that,
begin{align*}
R(1)=int_0^1 frac{tln t}{1+t}dt+int_0^1 frac{ln t}{1-t}dt
end{align*}
Perform integration by parts,
begin{align*}
I&=-int_0^1 frac{ln x}{1+x^2}lnleft(frac{1+x^2}{(1-x)^2}right)dx-2Big[R(x)arctan xBig]_0^1+2int_0^1frac{R(x)}{1+x^2}dx\
&=-int_0^1 frac{ln x}{1+x^2}lnleft(frac{1+x^2}{(1-x)^2}right)dx-frac{pi}{2}R(1)+2int_0^1 int_0^1 frac{x(1+tx)ln (tx)}{(1-tx)(1+t^2x^2)(1+x^2)}dtdx\
&=-int_0^1 frac{ln x}{1+x^2}lnleft(frac{1+x^2}{(1-x)^2}right)dx-frac{pi}{2}R(1)+int_0^1 ln xleft[frac{1}{1+x^2}lnleft(frac{1+t^2x^2}{(1-tx)^2}right)right]_{t=0}^{t=1} dx+\
&int_0^1 ln tleft[frac{1}{1+t^2}lnleft(frac{1+x^2}{(1-tx)^2}right)+frac{2arctan (tx)}{1-t^2}-frac{2tarctan x}{1+t^2}-frac{2tarctan x}{1-t^2}right]_{x=0}^{x=1} dt\
&=-frac{pi }{2}R(1)+ln 2int_0^1 frac{ln t}{1+t^2}dt-2int_0^1 frac{ln (1-t)ln t}{1+t^2}dt+2int_0^1 frac{ln tarctan t}{1-t^2}dt-\
&frac{pi}{2} int_0^1 frac{tln t}{1+t^2}dt-frac{pi}{2} int_0^1frac{tln t}{1-t^2} dt\
end{align*}
For $xin [0;1]$ define the function $S$ by,
begin{align*}
S(x)=int_0^x frac{ln t}{1-t^2}dt=int_0^1 frac{xln(tx)}{1-t^2x^2} dt
end{align*}
Perform integration by parts,
begin{align*}
int_0^1 frac{ln xarctan x}{1-x^2}dx&=Big[S(x)arctan xBig]_0^1-int_0^1 frac{S(x)}{1+x^2}dx\
&=frac{pi}{4}S(1)-int_0^1 int_0^1 frac{xln(tx)} {(1-t^2x^2)(1+x^2)} dtdx\
&=frac{pi}{4}S(1)-frac{1}{2}int_0^1 left[ frac{ln x}{1+x^2}lnleft(frac{1+tx}{1-tx} right)right]_{t=0}^{t=1} dx-\
&frac{1}{2}int_0^1 left[ frac{ln t}{1+t^2}lnleft(frac{1+x^2}{1-t^2x^2} right)right]_{x=0}^{x=1}dt\
&=frac{pi}{4}S(1)-frac{ln 2}{2}int_0^1 frac{ln t}{1+t^2}dt+int_0^1 frac{ln(1-x)ln x}{1+x^2}dx
end{align*}
Therefore,
begin{align*}I&=piint_0^1frac{2tln t}{t^4-1} dtend{align*}
Perform the change of variable $y=t^2$,
begin{align*}I&=frac{1}{2}pi int_0^1 frac{ln y}{y^2-1}dy\
&=frac{1}{2}pitimes frac{3}{4}zeta(2)\
&=frac{pi^3}{16}
end{align*}
$endgroup$
1
$begingroup$
That's impressive, thank you! I've seen you use this approach alot and it's quite useful, let me a few time to understand it's working better.
$endgroup$
– Zacky
Dec 25 '18 at 17:59
$begingroup$
Well done. (+1)
$endgroup$
– Mark Viola
Dec 26 '18 at 4:26
$begingroup$
Very nice solution and $to +1$
$endgroup$
– Claude Leibovici
Dec 26 '18 at 6:07
1
$begingroup$
I compute $int_0^1 F(t,x)ln t,dx$ and $int_0^1 F(t,x)ln x,dt$ and one can compute an antiderivative $U(t,x)$ of $F(t,x)$ wrt $x$, and on the other hand an antiderivative $V(t,x)$ of $F(t,x)$ wrt $t$.
$endgroup$
– FDP
Dec 26 '18 at 17:07
1
$begingroup$
@Zacky: remember in the double integrals you can choose to integrate wrt $x$ or wrt $t$. If there is a factor $ln x$ you don't want to integrate wrt $x$ first. If there is a factor $ln t$ you don't want to integrate wrt $t$ first. And, $ln(tx)=ln x +ln t$
$endgroup$
– FDP
Dec 26 '18 at 17:25
|
show 1 more comment
$begingroup$
Another approach,
Perform integration by parts,
begin{align*}
I&=int_0^1 frac{arctan x}{x}lnleft(frac{1+x^2}{(1-x)^2}right),dx\
&=Big[ln (x) lnleft(frac{1+x^2}{(1-x)^2}right)arctan xBig]_0^1 -int_0^1 frac{ln x}{1+x^2}lnleft(frac{1+x^2}{(1-x)^2}right)dx-int_0^1 frac{2(1+x)ln (x)arctan (x)}{(1-x)(1+x^2)}dx\
&=-int_0^1 frac{ln x}{1+x^2}lnleft(frac{1+x^2}{(1-x)^2}right)dx-2int_0^1 frac{(1+x)ln (x)arctan (x)}{(1-x)(1+x^2)}dx\
end{align*}
For $xin [0;1]$ define the function $R$ by,
begin{align*}
R(x)=int_0^x frac{(1+t)ln t}{(1-t)(1+t^2)}dt=int_0^1 frac{x(1+tx)ln (tx)}{(1-tx)(1+t^2x^2)}dt\
end{align*}
Observe that,
begin{align*}
R(1)=int_0^1 frac{tln t}{1+t}dt+int_0^1 frac{ln t}{1-t}dt
end{align*}
Perform integration by parts,
begin{align*}
I&=-int_0^1 frac{ln x}{1+x^2}lnleft(frac{1+x^2}{(1-x)^2}right)dx-2Big[R(x)arctan xBig]_0^1+2int_0^1frac{R(x)}{1+x^2}dx\
&=-int_0^1 frac{ln x}{1+x^2}lnleft(frac{1+x^2}{(1-x)^2}right)dx-frac{pi}{2}R(1)+2int_0^1 int_0^1 frac{x(1+tx)ln (tx)}{(1-tx)(1+t^2x^2)(1+x^2)}dtdx\
&=-int_0^1 frac{ln x}{1+x^2}lnleft(frac{1+x^2}{(1-x)^2}right)dx-frac{pi}{2}R(1)+int_0^1 ln xleft[frac{1}{1+x^2}lnleft(frac{1+t^2x^2}{(1-tx)^2}right)right]_{t=0}^{t=1} dx+\
&int_0^1 ln tleft[frac{1}{1+t^2}lnleft(frac{1+x^2}{(1-tx)^2}right)+frac{2arctan (tx)}{1-t^2}-frac{2tarctan x}{1+t^2}-frac{2tarctan x}{1-t^2}right]_{x=0}^{x=1} dt\
&=-frac{pi }{2}R(1)+ln 2int_0^1 frac{ln t}{1+t^2}dt-2int_0^1 frac{ln (1-t)ln t}{1+t^2}dt+2int_0^1 frac{ln tarctan t}{1-t^2}dt-\
&frac{pi}{2} int_0^1 frac{tln t}{1+t^2}dt-frac{pi}{2} int_0^1frac{tln t}{1-t^2} dt\
end{align*}
For $xin [0;1]$ define the function $S$ by,
begin{align*}
S(x)=int_0^x frac{ln t}{1-t^2}dt=int_0^1 frac{xln(tx)}{1-t^2x^2} dt
end{align*}
Perform integration by parts,
begin{align*}
int_0^1 frac{ln xarctan x}{1-x^2}dx&=Big[S(x)arctan xBig]_0^1-int_0^1 frac{S(x)}{1+x^2}dx\
&=frac{pi}{4}S(1)-int_0^1 int_0^1 frac{xln(tx)} {(1-t^2x^2)(1+x^2)} dtdx\
&=frac{pi}{4}S(1)-frac{1}{2}int_0^1 left[ frac{ln x}{1+x^2}lnleft(frac{1+tx}{1-tx} right)right]_{t=0}^{t=1} dx-\
&frac{1}{2}int_0^1 left[ frac{ln t}{1+t^2}lnleft(frac{1+x^2}{1-t^2x^2} right)right]_{x=0}^{x=1}dt\
&=frac{pi}{4}S(1)-frac{ln 2}{2}int_0^1 frac{ln t}{1+t^2}dt+int_0^1 frac{ln(1-x)ln x}{1+x^2}dx
end{align*}
Therefore,
begin{align*}I&=piint_0^1frac{2tln t}{t^4-1} dtend{align*}
Perform the change of variable $y=t^2$,
begin{align*}I&=frac{1}{2}pi int_0^1 frac{ln y}{y^2-1}dy\
&=frac{1}{2}pitimes frac{3}{4}zeta(2)\
&=frac{pi^3}{16}
end{align*}
$endgroup$
1
$begingroup$
That's impressive, thank you! I've seen you use this approach alot and it's quite useful, let me a few time to understand it's working better.
$endgroup$
– Zacky
Dec 25 '18 at 17:59
$begingroup$
Well done. (+1)
$endgroup$
– Mark Viola
Dec 26 '18 at 4:26
$begingroup$
Very nice solution and $to +1$
$endgroup$
– Claude Leibovici
Dec 26 '18 at 6:07
1
$begingroup$
I compute $int_0^1 F(t,x)ln t,dx$ and $int_0^1 F(t,x)ln x,dt$ and one can compute an antiderivative $U(t,x)$ of $F(t,x)$ wrt $x$, and on the other hand an antiderivative $V(t,x)$ of $F(t,x)$ wrt $t$.
$endgroup$
– FDP
Dec 26 '18 at 17:07
1
$begingroup$
@Zacky: remember in the double integrals you can choose to integrate wrt $x$ or wrt $t$. If there is a factor $ln x$ you don't want to integrate wrt $x$ first. If there is a factor $ln t$ you don't want to integrate wrt $t$ first. And, $ln(tx)=ln x +ln t$
$endgroup$
– FDP
Dec 26 '18 at 17:25
|
show 1 more comment
$begingroup$
Another approach,
Perform integration by parts,
begin{align*}
I&=int_0^1 frac{arctan x}{x}lnleft(frac{1+x^2}{(1-x)^2}right),dx\
&=Big[ln (x) lnleft(frac{1+x^2}{(1-x)^2}right)arctan xBig]_0^1 -int_0^1 frac{ln x}{1+x^2}lnleft(frac{1+x^2}{(1-x)^2}right)dx-int_0^1 frac{2(1+x)ln (x)arctan (x)}{(1-x)(1+x^2)}dx\
&=-int_0^1 frac{ln x}{1+x^2}lnleft(frac{1+x^2}{(1-x)^2}right)dx-2int_0^1 frac{(1+x)ln (x)arctan (x)}{(1-x)(1+x^2)}dx\
end{align*}
For $xin [0;1]$ define the function $R$ by,
begin{align*}
R(x)=int_0^x frac{(1+t)ln t}{(1-t)(1+t^2)}dt=int_0^1 frac{x(1+tx)ln (tx)}{(1-tx)(1+t^2x^2)}dt\
end{align*}
Observe that,
begin{align*}
R(1)=int_0^1 frac{tln t}{1+t}dt+int_0^1 frac{ln t}{1-t}dt
end{align*}
Perform integration by parts,
begin{align*}
I&=-int_0^1 frac{ln x}{1+x^2}lnleft(frac{1+x^2}{(1-x)^2}right)dx-2Big[R(x)arctan xBig]_0^1+2int_0^1frac{R(x)}{1+x^2}dx\
&=-int_0^1 frac{ln x}{1+x^2}lnleft(frac{1+x^2}{(1-x)^2}right)dx-frac{pi}{2}R(1)+2int_0^1 int_0^1 frac{x(1+tx)ln (tx)}{(1-tx)(1+t^2x^2)(1+x^2)}dtdx\
&=-int_0^1 frac{ln x}{1+x^2}lnleft(frac{1+x^2}{(1-x)^2}right)dx-frac{pi}{2}R(1)+int_0^1 ln xleft[frac{1}{1+x^2}lnleft(frac{1+t^2x^2}{(1-tx)^2}right)right]_{t=0}^{t=1} dx+\
&int_0^1 ln tleft[frac{1}{1+t^2}lnleft(frac{1+x^2}{(1-tx)^2}right)+frac{2arctan (tx)}{1-t^2}-frac{2tarctan x}{1+t^2}-frac{2tarctan x}{1-t^2}right]_{x=0}^{x=1} dt\
&=-frac{pi }{2}R(1)+ln 2int_0^1 frac{ln t}{1+t^2}dt-2int_0^1 frac{ln (1-t)ln t}{1+t^2}dt+2int_0^1 frac{ln tarctan t}{1-t^2}dt-\
&frac{pi}{2} int_0^1 frac{tln t}{1+t^2}dt-frac{pi}{2} int_0^1frac{tln t}{1-t^2} dt\
end{align*}
For $xin [0;1]$ define the function $S$ by,
begin{align*}
S(x)=int_0^x frac{ln t}{1-t^2}dt=int_0^1 frac{xln(tx)}{1-t^2x^2} dt
end{align*}
Perform integration by parts,
begin{align*}
int_0^1 frac{ln xarctan x}{1-x^2}dx&=Big[S(x)arctan xBig]_0^1-int_0^1 frac{S(x)}{1+x^2}dx\
&=frac{pi}{4}S(1)-int_0^1 int_0^1 frac{xln(tx)} {(1-t^2x^2)(1+x^2)} dtdx\
&=frac{pi}{4}S(1)-frac{1}{2}int_0^1 left[ frac{ln x}{1+x^2}lnleft(frac{1+tx}{1-tx} right)right]_{t=0}^{t=1} dx-\
&frac{1}{2}int_0^1 left[ frac{ln t}{1+t^2}lnleft(frac{1+x^2}{1-t^2x^2} right)right]_{x=0}^{x=1}dt\
&=frac{pi}{4}S(1)-frac{ln 2}{2}int_0^1 frac{ln t}{1+t^2}dt+int_0^1 frac{ln(1-x)ln x}{1+x^2}dx
end{align*}
Therefore,
begin{align*}I&=piint_0^1frac{2tln t}{t^4-1} dtend{align*}
Perform the change of variable $y=t^2$,
begin{align*}I&=frac{1}{2}pi int_0^1 frac{ln y}{y^2-1}dy\
&=frac{1}{2}pitimes frac{3}{4}zeta(2)\
&=frac{pi^3}{16}
end{align*}
$endgroup$
Another approach,
Perform integration by parts,
begin{align*}
I&=int_0^1 frac{arctan x}{x}lnleft(frac{1+x^2}{(1-x)^2}right),dx\
&=Big[ln (x) lnleft(frac{1+x^2}{(1-x)^2}right)arctan xBig]_0^1 -int_0^1 frac{ln x}{1+x^2}lnleft(frac{1+x^2}{(1-x)^2}right)dx-int_0^1 frac{2(1+x)ln (x)arctan (x)}{(1-x)(1+x^2)}dx\
&=-int_0^1 frac{ln x}{1+x^2}lnleft(frac{1+x^2}{(1-x)^2}right)dx-2int_0^1 frac{(1+x)ln (x)arctan (x)}{(1-x)(1+x^2)}dx\
end{align*}
For $xin [0;1]$ define the function $R$ by,
begin{align*}
R(x)=int_0^x frac{(1+t)ln t}{(1-t)(1+t^2)}dt=int_0^1 frac{x(1+tx)ln (tx)}{(1-tx)(1+t^2x^2)}dt\
end{align*}
Observe that,
begin{align*}
R(1)=int_0^1 frac{tln t}{1+t}dt+int_0^1 frac{ln t}{1-t}dt
end{align*}
Perform integration by parts,
begin{align*}
I&=-int_0^1 frac{ln x}{1+x^2}lnleft(frac{1+x^2}{(1-x)^2}right)dx-2Big[R(x)arctan xBig]_0^1+2int_0^1frac{R(x)}{1+x^2}dx\
&=-int_0^1 frac{ln x}{1+x^2}lnleft(frac{1+x^2}{(1-x)^2}right)dx-frac{pi}{2}R(1)+2int_0^1 int_0^1 frac{x(1+tx)ln (tx)}{(1-tx)(1+t^2x^2)(1+x^2)}dtdx\
&=-int_0^1 frac{ln x}{1+x^2}lnleft(frac{1+x^2}{(1-x)^2}right)dx-frac{pi}{2}R(1)+int_0^1 ln xleft[frac{1}{1+x^2}lnleft(frac{1+t^2x^2}{(1-tx)^2}right)right]_{t=0}^{t=1} dx+\
&int_0^1 ln tleft[frac{1}{1+t^2}lnleft(frac{1+x^2}{(1-tx)^2}right)+frac{2arctan (tx)}{1-t^2}-frac{2tarctan x}{1+t^2}-frac{2tarctan x}{1-t^2}right]_{x=0}^{x=1} dt\
&=-frac{pi }{2}R(1)+ln 2int_0^1 frac{ln t}{1+t^2}dt-2int_0^1 frac{ln (1-t)ln t}{1+t^2}dt+2int_0^1 frac{ln tarctan t}{1-t^2}dt-\
&frac{pi}{2} int_0^1 frac{tln t}{1+t^2}dt-frac{pi}{2} int_0^1frac{tln t}{1-t^2} dt\
end{align*}
For $xin [0;1]$ define the function $S$ by,
begin{align*}
S(x)=int_0^x frac{ln t}{1-t^2}dt=int_0^1 frac{xln(tx)}{1-t^2x^2} dt
end{align*}
Perform integration by parts,
begin{align*}
int_0^1 frac{ln xarctan x}{1-x^2}dx&=Big[S(x)arctan xBig]_0^1-int_0^1 frac{S(x)}{1+x^2}dx\
&=frac{pi}{4}S(1)-int_0^1 int_0^1 frac{xln(tx)} {(1-t^2x^2)(1+x^2)} dtdx\
&=frac{pi}{4}S(1)-frac{1}{2}int_0^1 left[ frac{ln x}{1+x^2}lnleft(frac{1+tx}{1-tx} right)right]_{t=0}^{t=1} dx-\
&frac{1}{2}int_0^1 left[ frac{ln t}{1+t^2}lnleft(frac{1+x^2}{1-t^2x^2} right)right]_{x=0}^{x=1}dt\
&=frac{pi}{4}S(1)-frac{ln 2}{2}int_0^1 frac{ln t}{1+t^2}dt+int_0^1 frac{ln(1-x)ln x}{1+x^2}dx
end{align*}
Therefore,
begin{align*}I&=piint_0^1frac{2tln t}{t^4-1} dtend{align*}
Perform the change of variable $y=t^2$,
begin{align*}I&=frac{1}{2}pi int_0^1 frac{ln y}{y^2-1}dy\
&=frac{1}{2}pitimes frac{3}{4}zeta(2)\
&=frac{pi^3}{16}
end{align*}
edited Dec 26 '18 at 17:22
answered Dec 25 '18 at 17:07
FDPFDP
5,47211525
5,47211525
1
$begingroup$
That's impressive, thank you! I've seen you use this approach alot and it's quite useful, let me a few time to understand it's working better.
$endgroup$
– Zacky
Dec 25 '18 at 17:59
$begingroup$
Well done. (+1)
$endgroup$
– Mark Viola
Dec 26 '18 at 4:26
$begingroup$
Very nice solution and $to +1$
$endgroup$
– Claude Leibovici
Dec 26 '18 at 6:07
1
$begingroup$
I compute $int_0^1 F(t,x)ln t,dx$ and $int_0^1 F(t,x)ln x,dt$ and one can compute an antiderivative $U(t,x)$ of $F(t,x)$ wrt $x$, and on the other hand an antiderivative $V(t,x)$ of $F(t,x)$ wrt $t$.
$endgroup$
– FDP
Dec 26 '18 at 17:07
1
$begingroup$
@Zacky: remember in the double integrals you can choose to integrate wrt $x$ or wrt $t$. If there is a factor $ln x$ you don't want to integrate wrt $x$ first. If there is a factor $ln t$ you don't want to integrate wrt $t$ first. And, $ln(tx)=ln x +ln t$
$endgroup$
– FDP
Dec 26 '18 at 17:25
|
show 1 more comment
1
$begingroup$
That's impressive, thank you! I've seen you use this approach alot and it's quite useful, let me a few time to understand it's working better.
$endgroup$
– Zacky
Dec 25 '18 at 17:59
$begingroup$
Well done. (+1)
$endgroup$
– Mark Viola
Dec 26 '18 at 4:26
$begingroup$
Very nice solution and $to +1$
$endgroup$
– Claude Leibovici
Dec 26 '18 at 6:07
1
$begingroup$
I compute $int_0^1 F(t,x)ln t,dx$ and $int_0^1 F(t,x)ln x,dt$ and one can compute an antiderivative $U(t,x)$ of $F(t,x)$ wrt $x$, and on the other hand an antiderivative $V(t,x)$ of $F(t,x)$ wrt $t$.
$endgroup$
– FDP
Dec 26 '18 at 17:07
1
$begingroup$
@Zacky: remember in the double integrals you can choose to integrate wrt $x$ or wrt $t$. If there is a factor $ln x$ you don't want to integrate wrt $x$ first. If there is a factor $ln t$ you don't want to integrate wrt $t$ first. And, $ln(tx)=ln x +ln t$
$endgroup$
– FDP
Dec 26 '18 at 17:25
1
1
$begingroup$
That's impressive, thank you! I've seen you use this approach alot and it's quite useful, let me a few time to understand it's working better.
$endgroup$
– Zacky
Dec 25 '18 at 17:59
$begingroup$
That's impressive, thank you! I've seen you use this approach alot and it's quite useful, let me a few time to understand it's working better.
$endgroup$
– Zacky
Dec 25 '18 at 17:59
$begingroup$
Well done. (+1)
$endgroup$
– Mark Viola
Dec 26 '18 at 4:26
$begingroup$
Well done. (+1)
$endgroup$
– Mark Viola
Dec 26 '18 at 4:26
$begingroup$
Very nice solution and $to +1$
$endgroup$
– Claude Leibovici
Dec 26 '18 at 6:07
$begingroup$
Very nice solution and $to +1$
$endgroup$
– Claude Leibovici
Dec 26 '18 at 6:07
1
1
$begingroup$
I compute $int_0^1 F(t,x)ln t,dx$ and $int_0^1 F(t,x)ln x,dt$ and one can compute an antiderivative $U(t,x)$ of $F(t,x)$ wrt $x$, and on the other hand an antiderivative $V(t,x)$ of $F(t,x)$ wrt $t$.
$endgroup$
– FDP
Dec 26 '18 at 17:07
$begingroup$
I compute $int_0^1 F(t,x)ln t,dx$ and $int_0^1 F(t,x)ln x,dt$ and one can compute an antiderivative $U(t,x)$ of $F(t,x)$ wrt $x$, and on the other hand an antiderivative $V(t,x)$ of $F(t,x)$ wrt $t$.
$endgroup$
– FDP
Dec 26 '18 at 17:07
1
1
$begingroup$
@Zacky: remember in the double integrals you can choose to integrate wrt $x$ or wrt $t$. If there is a factor $ln x$ you don't want to integrate wrt $x$ first. If there is a factor $ln t$ you don't want to integrate wrt $t$ first. And, $ln(tx)=ln x +ln t$
$endgroup$
– FDP
Dec 26 '18 at 17:25
$begingroup$
@Zacky: remember in the double integrals you can choose to integrate wrt $x$ or wrt $t$. If there is a factor $ln x$ you don't want to integrate wrt $x$ first. If there is a factor $ln t$ you don't want to integrate wrt $t$ first. And, $ln(tx)=ln x +ln t$
$endgroup$
– FDP
Dec 26 '18 at 17:25
|
show 1 more comment
$begingroup$
Put
begin{equation*}
I=int_{0}^1dfrac{arctan x}{x}lnleft(dfrac{1+x^2}{(1-x)^2}right), mathrm{d}x.
end{equation*}
Via the substitution $ x=dfrac{z}{z+1}$ we get
begin{equation*}
I = int_{0}^{infty}dfrac{arctan frac{z}{z+1}ln(2z^2+2z+1)}{z^2+z}, mathrm{d}z.
end{equation*}
Put
begin{equation*}
log z=ln|z|+iarg z, quad -pi<arg z <pi.
end{equation*}
Then
begin{equation*}
arctan frac{z}{z+1}ln(2z^2+2z+1) = text{Im}left(log^2(1+z+iz)right).
end{equation*}
Consequently
begin{equation*}
I = text{Im}left(int_{0}^{infty}dfrac{log^2(1+z+iz)}{z^2+z}right)mathrm{d}z.
end{equation*}
However, $ log(z) $ is an analytic function in $ text{Re} z>0 $. According to Cauchys integral theorem we get the same value if we integrate along the curve with the parametrization $ z=(1-i)s, s>0 $.
begin{gather*}
I = text{Im}left(int_{0}^{infty}dfrac{ln^2(2s+1)}{s(s+1-is)}, mathrm{d}sright) = int_{0}^{infty}dfrac{ln^2(2s+1)}{2s^2+2s+1}, mathrm{d}s = \[2ex] int_{0}^{infty}dfrac{2ln^2(2s+1)}{(2s+1)^2+1}, mathrm{d}s = [t=2s+1] = \[2ex] int_{1}^{infty}dfrac{ln^2(t)}{t^2+1}, mathrm{d}t =[u= 1/t] = int_{0}^{1}dfrac{ln^2(u)}{u^2+1}, mathrm{d}u.
end{gather*}
Thus
begin{equation*}
2I = int_{0}^{infty}dfrac{ln^2(u)}{u^2+1}, mathrm{d}u
end{equation*}
In order to evaluate this integral we integrate $ dfrac{log^3(z)}{z^2+1} $ along a keyhole contour and use residue calculus.
In this case $ log z =ln |z|+iarg z, quad 0<arg z < 2pi $.
We get
begin{equation*}
I = dfrac{pi^3}{16}.
end{equation*}
$endgroup$
add a comment |
$begingroup$
Put
begin{equation*}
I=int_{0}^1dfrac{arctan x}{x}lnleft(dfrac{1+x^2}{(1-x)^2}right), mathrm{d}x.
end{equation*}
Via the substitution $ x=dfrac{z}{z+1}$ we get
begin{equation*}
I = int_{0}^{infty}dfrac{arctan frac{z}{z+1}ln(2z^2+2z+1)}{z^2+z}, mathrm{d}z.
end{equation*}
Put
begin{equation*}
log z=ln|z|+iarg z, quad -pi<arg z <pi.
end{equation*}
Then
begin{equation*}
arctan frac{z}{z+1}ln(2z^2+2z+1) = text{Im}left(log^2(1+z+iz)right).
end{equation*}
Consequently
begin{equation*}
I = text{Im}left(int_{0}^{infty}dfrac{log^2(1+z+iz)}{z^2+z}right)mathrm{d}z.
end{equation*}
However, $ log(z) $ is an analytic function in $ text{Re} z>0 $. According to Cauchys integral theorem we get the same value if we integrate along the curve with the parametrization $ z=(1-i)s, s>0 $.
begin{gather*}
I = text{Im}left(int_{0}^{infty}dfrac{ln^2(2s+1)}{s(s+1-is)}, mathrm{d}sright) = int_{0}^{infty}dfrac{ln^2(2s+1)}{2s^2+2s+1}, mathrm{d}s = \[2ex] int_{0}^{infty}dfrac{2ln^2(2s+1)}{(2s+1)^2+1}, mathrm{d}s = [t=2s+1] = \[2ex] int_{1}^{infty}dfrac{ln^2(t)}{t^2+1}, mathrm{d}t =[u= 1/t] = int_{0}^{1}dfrac{ln^2(u)}{u^2+1}, mathrm{d}u.
end{gather*}
Thus
begin{equation*}
2I = int_{0}^{infty}dfrac{ln^2(u)}{u^2+1}, mathrm{d}u
end{equation*}
In order to evaluate this integral we integrate $ dfrac{log^3(z)}{z^2+1} $ along a keyhole contour and use residue calculus.
In this case $ log z =ln |z|+iarg z, quad 0<arg z < 2pi $.
We get
begin{equation*}
I = dfrac{pi^3}{16}.
end{equation*}
$endgroup$
add a comment |
$begingroup$
Put
begin{equation*}
I=int_{0}^1dfrac{arctan x}{x}lnleft(dfrac{1+x^2}{(1-x)^2}right), mathrm{d}x.
end{equation*}
Via the substitution $ x=dfrac{z}{z+1}$ we get
begin{equation*}
I = int_{0}^{infty}dfrac{arctan frac{z}{z+1}ln(2z^2+2z+1)}{z^2+z}, mathrm{d}z.
end{equation*}
Put
begin{equation*}
log z=ln|z|+iarg z, quad -pi<arg z <pi.
end{equation*}
Then
begin{equation*}
arctan frac{z}{z+1}ln(2z^2+2z+1) = text{Im}left(log^2(1+z+iz)right).
end{equation*}
Consequently
begin{equation*}
I = text{Im}left(int_{0}^{infty}dfrac{log^2(1+z+iz)}{z^2+z}right)mathrm{d}z.
end{equation*}
However, $ log(z) $ is an analytic function in $ text{Re} z>0 $. According to Cauchys integral theorem we get the same value if we integrate along the curve with the parametrization $ z=(1-i)s, s>0 $.
begin{gather*}
I = text{Im}left(int_{0}^{infty}dfrac{ln^2(2s+1)}{s(s+1-is)}, mathrm{d}sright) = int_{0}^{infty}dfrac{ln^2(2s+1)}{2s^2+2s+1}, mathrm{d}s = \[2ex] int_{0}^{infty}dfrac{2ln^2(2s+1)}{(2s+1)^2+1}, mathrm{d}s = [t=2s+1] = \[2ex] int_{1}^{infty}dfrac{ln^2(t)}{t^2+1}, mathrm{d}t =[u= 1/t] = int_{0}^{1}dfrac{ln^2(u)}{u^2+1}, mathrm{d}u.
end{gather*}
Thus
begin{equation*}
2I = int_{0}^{infty}dfrac{ln^2(u)}{u^2+1}, mathrm{d}u
end{equation*}
In order to evaluate this integral we integrate $ dfrac{log^3(z)}{z^2+1} $ along a keyhole contour and use residue calculus.
In this case $ log z =ln |z|+iarg z, quad 0<arg z < 2pi $.
We get
begin{equation*}
I = dfrac{pi^3}{16}.
end{equation*}
$endgroup$
Put
begin{equation*}
I=int_{0}^1dfrac{arctan x}{x}lnleft(dfrac{1+x^2}{(1-x)^2}right), mathrm{d}x.
end{equation*}
Via the substitution $ x=dfrac{z}{z+1}$ we get
begin{equation*}
I = int_{0}^{infty}dfrac{arctan frac{z}{z+1}ln(2z^2+2z+1)}{z^2+z}, mathrm{d}z.
end{equation*}
Put
begin{equation*}
log z=ln|z|+iarg z, quad -pi<arg z <pi.
end{equation*}
Then
begin{equation*}
arctan frac{z}{z+1}ln(2z^2+2z+1) = text{Im}left(log^2(1+z+iz)right).
end{equation*}
Consequently
begin{equation*}
I = text{Im}left(int_{0}^{infty}dfrac{log^2(1+z+iz)}{z^2+z}right)mathrm{d}z.
end{equation*}
However, $ log(z) $ is an analytic function in $ text{Re} z>0 $. According to Cauchys integral theorem we get the same value if we integrate along the curve with the parametrization $ z=(1-i)s, s>0 $.
begin{gather*}
I = text{Im}left(int_{0}^{infty}dfrac{ln^2(2s+1)}{s(s+1-is)}, mathrm{d}sright) = int_{0}^{infty}dfrac{ln^2(2s+1)}{2s^2+2s+1}, mathrm{d}s = \[2ex] int_{0}^{infty}dfrac{2ln^2(2s+1)}{(2s+1)^2+1}, mathrm{d}s = [t=2s+1] = \[2ex] int_{1}^{infty}dfrac{ln^2(t)}{t^2+1}, mathrm{d}t =[u= 1/t] = int_{0}^{1}dfrac{ln^2(u)}{u^2+1}, mathrm{d}u.
end{gather*}
Thus
begin{equation*}
2I = int_{0}^{infty}dfrac{ln^2(u)}{u^2+1}, mathrm{d}u
end{equation*}
In order to evaluate this integral we integrate $ dfrac{log^3(z)}{z^2+1} $ along a keyhole contour and use residue calculus.
In this case $ log z =ln |z|+iarg z, quad 0<arg z < 2pi $.
We get
begin{equation*}
I = dfrac{pi^3}{16}.
end{equation*}
answered Dec 29 '18 at 23:03
JanGJanG
2,802514
2,802514
add a comment |
add a comment |
Thanks for contributing an answer to Mathematics Stack Exchange!
- Please be sure to answer the question. Provide details and share your research!
But avoid …
- Asking for help, clarification, or responding to other answers.
- Making statements based on opinion; back them up with references or personal experience.
Use MathJax to format equations. MathJax reference.
To learn more, see our tips on writing great answers.
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
StackExchange.ready(
function () {
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fmath.stackexchange.com%2fquestions%2f3050696%2fproving-that-int-01-frac-arctan-xx-ln-left-frac1x21-x2-rightd%23new-answer', 'question_page');
}
);
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
3
$begingroup$
I was able to break it down to an evaluation of harmonic sums $$I=frac{3pi^3}{32}-sum_{n=0}^{infty}frac{frac12left[H_{n/2}-H_{(n-1)/2}right]+frac14left[H_{n+1/4}-H_{n-1/4}right]}{(2n+1)^2}$$ the latter sum should equal $pi^3/32$ which seems to work out numerically but honestly speaking I am lost from hereon. Using the well-known result $beta(3)=pi^3/32$ one could conjecture that the combination of harmonic sums has to come out equal to $(-1)^n/(2n+1)$ in order to complete the representation of $beta(3)$.
$endgroup$
– mrtaurho
Dec 24 '18 at 1:48
1
$begingroup$
I would say this solution here is quite impressive and convincing.
$endgroup$
– mrtaurho
Dec 25 '18 at 15:38