0 1 x-x dx


By Pooya Farshim (P2572) on Saturday, June 3, 2000 - 08:05 pm :

Prove that:


0 1 x-x dx= i=1 i-i
I tried the trapezium rule but got stuck in the mid-way!
By Michael Doré (P904) on Wednesday, June 7, 2000 - 02:03 pm :

How about this?

Let I(a,b)= xa ln(x )b dx

(By the way all integrals here are with respect to x and between 0 and 1.)

Integration by parts reveals that:

I(a,b)=( xa+1 ln(x )b )/(a+1)-b/(a+1)I(a,b-1)

Provided a+1b the first term on the right hand side will be zero after the limits are applied, so we can then say:

I(a,b)=-b/(a+1)I(a,b-1)

By successive application of this formula we can evaluate I(n,n) (for +ve integral n) as:

I(n,n)=(-n)(-n+1)(-n+2)...(-2)(-1)/(n+1 )n ×I(n,0)

Clearly I(n,0)=1/(n+1)

So I(n,n)=(-1 )n n!/(n+1 )n+1

Now back to the problem.

x-x dx can be written as e-xlnx dx which can now be expanded out using the infinite series. We get

[1-xlnx+(xlnx )2 /2!-(xlnx )3 /3!+...]dx

=I(0,0)-I(1,1)+I(2,2)/2!-I(3,3)/3!+...

Now the factorials will cancel with the factorials contained within the I(n,n)s. The (-1 )n will disappear and we are left with:

1-1 + 2-2 + 3-3 +...


as required.

Yours,

Michael


By Pooya Farshim (P2572) on Wednesday, June 14, 2000 - 09:33 pm :

Thanks Michael