If the value of the integral
0
∫
∞
x
−
1
x
n
−
1
d
x
=
−
f
(
n
)
a
where
Find the value of a + ln ( c ) .
■ This is a part of Hot Integrals
This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try
refreshing the page, (b) enabling javascript if it is disabled on your browser and,
finally, (c)
loading the
non-javascript version of this page
. We're sorry about the hassle.
Good my friend ! I have a short solution which i will post after sometime.
Log in to reply
Why can't we have a = 2 π and c = 2 2 − 2 ?
Log in to reply
No i dont think so , while integrating this , one can get 2 π in the numerator.
This question can be done via complex integration.
Nice problem, seriously.
∫ 0 ∞ x − 1 x n − 1 d x
is our problem. Well, I will do it using contour integration.
What we can see? A clear pole at z = 1 (we will just change all the x to z since we are in Argand plane now) and there is a branch point as well z = 0 . That were the singularities and now we have to play with them.
Consider the contour as a full circle around the origin with a branch cut towards the positive real axis, in short, a "keyhole contour".
So, C = Γ + L 1 + L 2 + γ
Γ represents the "outside" of C with radius M( M ⇒ ∞ )
γ represents the small circle around the origin with radius ϵ ( ⇒ 0 )
L 1 represents the straight line extending from the small circle over the +ve real axis.
L 2 represents the straight line extending from the small circle below the +ve real axis.
We can see that
∫ Γ z − 1 z n − 1 ⇒ 0
∫ γ z − 1 z n − 1 ⇒ 0
*I am not giving the details of the above integrals but they can easily be seen.
And surely, we'll then evaluate the rest.
We can write z n − 1 = e ( n − 1 ) l n ( z ) . Why did we do that?
∫ L 1 z − 1 z n − 1 = ∫ 0 ∞ z − 1 e ( n − 1 ) l n ( z )
as L 1 is a straight line.
Similarly,
∫ L 1 z − 1 z n − 1 = − ∫ 0 ∞ z − 1 e ( n − 1 ) ( l n ( z ( e 2 i π ) )
We did that for this. Now, it becomes easy for us to get the answer.
Why did the minus sign come in front? L 2 is below so, the denominator would change as well.
Why did that e 2 i π come? Well, because that is the "phase difference" between the two.
Finally, we know that
C = Γ + L 1 + L 2 + γ
Hence,
∫ C z − 1 z n − 1 = ( 1 − e 2 ( n − 1 ) i π ) ( ∫ 0 ∞ z − 1 z n − 1 )
Now, the line integral around the curve C can be found using Residue theorem and comes out to be 2 i π e 2 i π ( n − 1 )
Therefore,
∫ 0 ∞ z − 1 z n − 1 = ( 1 − e 2 ( n − 1 ) i π ) 2 i π e 2 i π ( n − 1 )
which can be simplified to
= − t a n ( n π ) π − i π
Now, I don't how to get rid of that extra i π and it was the reason of I first reporting it. But still, we can see that we want a real value and consider the first. Help me here!
@Aman Raimann Where am I wrong? I didn't get it. Can you please help?
you forgot to take a pole at z=0
Good work @Kartik Sharma !
Please Can you tell me how I can learn contour integration? Still I don't know anything about complex analysis.. Can you suggest some books ?
Log in to reply
Well, I would suggest you to look out for it on Google and there you will find a link to MSE(Math StackExchange). There, you can get a better overview of all the available books. Well, for a beginner(as I'm too), I would suggest to take up some lectures notes of any university(as you like). They are usually a good read and I too learnt from them. Good luck! However, this is not good work since I am not able to get the correct answer, I am getting my answer as a complex number.
Can u tell me how did u find out line integral . its wrong ! because using residue theorem you have considered the singularity at z=1 which is outside the countour.
Log in to reply
The contour is a circle of infinite radius. I didn't get what you are saying. A branch cut doesn't make the singularity to go out of the "complete contour".
Problem Loading...
Note Loading...
Set Loading...
Our Integral can be written as:
I = ∫ 0 ∞ x − 1 x n − 1 d x = ∫ 0 1 − x − 1 x n − 1 d x + ∫ 1 + ∞ x − 1 x n − 1 d x
Use the substitution x → x 1 in the second integral, to get:
I = − ∫ 0 1 − 1 − x x n − 1 d x + ∫ 0 1 − 1 − x x − n d x
Now as long as the integral limits are the same, and bounded between 0 and 1 , we can rewrite I as:
I = − ∫ 0 1 − m = 0 ∑ ∞ ( x m + n − 1 − x m − n ) d x
Integrating:
I = − m = 0 ∑ ∞ ( m + n 1 + m − n + 1 1 )
= m = 1 ∑ ∞ m + ( 1 − n ) 1 − m = 1 ∑ ∞ m + n 1 − n 1 + 1 − n 1
Now we are going to show that :
m = 1 ∑ ∞ m + ( 1 − n ) 1 − m = 1 ∑ ∞ m + n 1 = H n − H 1 − n
Where H k is the k t h Harmonic Number, Generalized to the positive non-integral arguments.
Recall that for any real positive x :
H x = x k = 1 ∑ ∞ k ( k + x ) 1
= k = 1 ∑ ∞ k 1 − k = 1 ∑ ∞ k + x 1
Therefore:
H n − H 1 − n = [ k = 1 ∑ ∞ k 1 − k = 1 ∑ ∞ k + n 1 ] − [ k = 1 ∑ ∞ k 1 − k = 1 ∑ ∞ k + ( 1 − n ) 1 ]
= k = 1 ∑ ∞ k + ( 1 − n ) 1 − k = 1 ∑ ∞ k + n 1
So rewrite I :
I = H n − H 1 − n − n 1 + 1 − n 1
∙ P r e p o s i t i o n 1 :
For 0 < r < 1 , we have the following Reflection Relation:
H 1 − r − H r = π cot ( π r ) − r 1 + 1 − r 1
To prove this, we need other prepositions:
∙ P r e p o s i t i o n 2 :
For any positive real y :
∫ 0 y H x d x = y γ + ln ( y ! )
Where γ is the Euler-Mascheroni Constant .
P r o o f :
Start with the definition of the harmonic number, and integrate both sides :
∫ 0 y H x d x = ∫ 0 y x k = 1 ∑ ∞ k ( k + x ) 1 d x
(just basic integration and evaluation):
= i → ∞ lim k = 1 ∑ i ( k y − ln ( y + k ) + ln k )
= i → ∞ lim ( y H i − ln ( y ! ( y + i ) ! ) + ln i ! )
= i → ∞ lim ( y H i + ( − y ln i + y ln i ) − ln ( y ! ( y + i ) ! ) + ln i ! )
= i → ∞ lim ( y ( H i − ln i ) + ln ( ( y + i ) ! i y i ! ) + ln y ! )
Now, Recall the Definition of the Euler-Mascheroni Constant. Then notice that the red term will vanish using Stirling's Formula. So:
∫ 0 y H x d x = y γ + ln y !
∙ P r e p o s i t i o n 3 :
For any n > 0 :
ψ ( n ) = H n − n 1 − γ
Where : ψ ( n ) is the Digamma Function , and γ is the Euler-Mascheroni Constant .
P r o o f :
Using the identity #2 :
∫ 0 y H x d x = y γ + ln Γ ( y + 1 )
Differentiate w.r.t y Then Recall the definition of the Digamma function :
H y = γ + ψ ( y + 1 )
Therefore : ( y → ( n − 1 ) )
ψ ( n ) = H n − 1 − γ = H n − n 1 − γ
Now, we can prove our first preposition.
Recall the Well Known Euler's Reflection Formula for the Gamma Function:
Γ ( z ) Γ ( 1 − z ) = π csc ( π z )
Take the Natural Logarithm of both Sides:
ln Γ ( z ) + ln Γ ( 1 − z ) = ln π − ln sin ( π z )
Differentiate Both Sides:
ψ ( z ) − ψ ( 1 − z ) = − π cot ( π z )
Now use the Identity #3 :
H z − z 1 − H 1 − z + 1 − z 1 = − π cot ( π z )
Rearranging, the reflection relation is thus proved:
H 1 − r − H r = π cot ( π r ) − r 1 + 1 − r 1
Now evaluate this relation back in the equation of I :
I = − tan ( π n ) π
The rest work is pretty easy, use tan 8 π = 2 − 1 , and the final answer becomes: π + ln ( 2 − 1 ) .