5
$\begingroup$

Let $I=\int_{0}^{1}\sqrt{1-x^{\pi}}dx$ If we use $u=x^{\pi}$ then we get $dx=\frac{du}{\pi^{\frac{\pi-1}{\pi}}}$ are there any better identites or substitution that can make integral easier to compute?

$\endgroup$
2
  • $\begingroup$ This integral has no close form expression. $\endgroup$
    – NN2
    Feb 8 '21 at 10:25
  • 3
    $\begingroup$ The indefinite integral, does not have a closed form, to solve it you must surely use the hypergeometric function. On the definite integral, only one observation: In general, you can prove that $$\int_{0}^{1}\sqrt{1-x^{n}}\operatorname{dx}=\frac{\sqrt{\pi}\Gamma\left(1+\frac{1}{n}\right)}{2\Gamma\left( \frac{3}{2}+\frac{1}{n}\right)}, \quad n>0.$$ $\endgroup$
    – mathproof
    Feb 8 '21 at 10:29
9
$\begingroup$

[Note] : You can use the beta function :

  • Beta Function : $\displaystyle \mathfrak{B}(x,y):=\int_{0}^{1}(1-\mu)^{x-1}\mu^{y-1}\;\text{d}\mu=\frac{\Gamma(x)\Gamma(y)}{\Gamma(x+y)}$

Let $\mu:=x^{\pi}$. Then : \begin{align*} x^{\pi}=\mu&\implies\text{d}\mu=\pi x^{\pi -1}\\ \\ &\iff\text{d}x=\frac{\text{d}\mu}{\pi\cdot\mu^{\displaystyle\frac{\pi-1}{\pi}}} =\frac{1}{\pi}\mu^{\displaystyle\frac{1-\pi}{\pi}}\text{d}\mu \end{align*} Therefore : \begin{align*} I&=\frac{1}{\pi}\int_{0}^{1}\sqrt{1-\mu}\cdot\mu^{\displaystyle\frac{1-\pi}{\pi}}\text{d}\mu\\ \\ &=\frac{1}{\pi}\int_{0}^{1}(1-\mu)^{\frac{3}{2}-1}\mu^{\frac{1}{\pi}-1}\;\text{d}\mu\\ \\ &=\frac{1}{\pi}\mathfrak{B}\left(\frac{3}{2},\frac{1}{\pi}\right)\\ \\ &=\frac{1}{\pi}\frac{\Gamma\left(\frac{3}{2}\right)\Gamma\left(\frac{1}{\pi}\right)}{\Gamma\left(\frac{3}{2}+\frac{1}{\pi}\right)}\\ \\ &\approx 0.846941\cdots \end{align*}

$\endgroup$
3
$\begingroup$

Just for the fun of it

As said in comments and answers, the is no closed form expression for the antiderivative. However, we could have an infinite series representation of it since $$\sqrt{1-x^{\pi}}=\sqrt{1-x^{3}}\Bigg[1+x^3\sum_{n=1}^\infty \frac { 1} {2^n\, n! }\left(\frac{\log (x)}{x^3-1}\right)^n P_n(x)\,(\pi-3)^n\Bigg] $$ where the first $P_n(x)$ are $$\left( \begin{array}{cc} n & P_n(x) \\ 1 & 1 \\ 2 & x^3-2 \\ 3 & x^6-2 x^3+4 \\ 4 & x^9-4 x^6-4 x^3-8 \\ 5 & x^{12}-4 x^9+36 x^6+56 x^3+16 \\ 6 & x^{15}-6 x^{12}-76 x^9-496 x^6-336 x^3-32 \\ 7 & x^{18}-6 x^{15}+288 x^{12}+3088 x^9+5328 x^6+1632 x^3+64 \end{array} \right)$$

For any $n$, each term shows an explicit antiderivative in terms of hypergeometric function. For example, for $n=1$ $$\int \frac{ x^3 \log (x)}{ \sqrt{1-x^3}}\,dx=$$ $$-\frac{x^4}{16} \left(\, _3F_2\left(\frac{1}{2},\frac{4}{3},\frac{4}{3};\frac{7}{3},\frac{7}{3};x^3\right)- 4 \, _2F_1\left(\frac{1}{2},\frac{4}{3};\frac{7}{3};x^3\right) \log (x)\right)$$

Using only the elements given in this answer, computing the integral from $0$ to $t$ gives the following results

$$\left( \begin{array}{cccc} t & \text{approximation} & \text{solution} & \text{error} \\ 0.1 & 0.099991285257983 & 0.099991285257855 & 1.29 \times 10^{-13}\\ 0.2 & 0.199846061391490 & 0.199846061391227 & 2.63 \times 10^{-13}\\ 0.3 & 0.299172693249054 & 0.299172693248698 & 3.56 \times 10^{-13}\\ 0.4 & 0.397263323235192 & 0.397263323234765 & 4.26 \times 10^{-13}\\ 0.5 & 0.493045151618189 & 0.493045151617706 & 4.83 \times 10^{-13}\\ 0.6 & 0.584997454713703 & 0.584997454713173 & 5.30 \times 10^{-13}\\ 0.7 & 0.670989777542907 & 0.670989777542338 & 5.70 \times 10^{-13}\\ 0.8 & 0.747926304221105 & 0.747926304220506 & 6.00 \times 10^{-13}\\ 0.9 & 0.810765563739254 & 0.810765563738633 & 6.20 \times 10^{-13}\\ 1.0 & 0.846941809210854 & 0.846941809208595 & 2.26 \times 10^{-12} \end{array} \right)$$

Otherwise and much simpler (but more difficult to compute) $$I=\int\sqrt{1-x^{\pi}}\,dx=\frac{x \left(\pi \, _2F_1\left(\frac{1}{2},\frac{1}{\pi };1+\frac{1}{\pi };x^{\pi }\right)+2 \sqrt{1-x^{\pi }}\right)}{2+\pi }$$

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.