Sunday 14 March 2010

But WHY?


The image is a pumpkin carved a few Halloweens ago by Sonja L. One of my Stats/Calc students (and a really good Bassoonist, Bassooner, Bassoon-enough)

Given the day, you may figure out this one...but the question above is the real issue. This was in an article from L. Short and J.P. Melville of the Napier University in Edinburgh, Scotland (which is, at least partly, on Napier's old estate)...

Take a unit square. Now step by step, create a unit rectangle on its right side (this is another square), then create a unit rectangle (2 x 1/2) on the top. Now go back and create a unit rectangle on the right side again (this one will be 2/3 by 3/2) then back to create on on the top.. Continue this forever and the sequence of ratios of the length to height of the total rectangle formed will converge to a Limit... find it...



and then explain WHY that is the limit... (blatant confession... I have no idea why this happens geometrically, and don't know anyone who does...)

Some more stuff from an older post on Pi I wrote a while back,,, after all, today is the day for it....
---------------------------
I know there are lots of infinite products that are equal to pi, and always thought
Wallis' expansion for pi/2 was beautiful,



And of course it is not as pretty, at least not to me, but since it was the first infinite product in math, I feel compelled to mention that Viete gave one for the reciprocal, 2/pi:

The question about this one for the really clever student is what does it have to do with the old calculus limit of Sin(theta)/theta... or the half angle formulas?

And Leibniz (you remember, the guy who invented Calculus if you DON'T live in England, just Kidding folks, it was Newton all the way, good job Ike) wrote one that Clifford Pickover calls "eye candy for pi";
pi/4 = 1 - 1/3 + 1/5 - 1/7+ 1/9....

except, it takes forever to converge.... the sum of the first 250 terms is not accurate to the second decimal place...

Ok, but the point of all this........

But today I came across one I had never seen, from the master of us all, Euler. Euler, it seems wrote pi/2 as an infinite product of fractions in which the numerators were all prime and the denominators were all even numbers excluding multiples of four. What appears from what I see is that each denominator is one more or less than the prime in the numerator, but always avoiding the one which would be a multiple of four... (Ok, now how do you write that as in product notation??).

pi/2 = 3/2 ( 5/6)(7/6)(11/10) (13/14)(17/18)

I looked for this a little and could not come up with a reference. If anyone knows where Euler wrote this, please advise.

Franz Gnaedinger who wrote the post where I picked this up, also pointed out that
"The analogous infinite product using all odd numbers
in the numerator seems to approximate the natural
logarithm of 2:

ln2 = 1/2 x 3/2 x 5/6 x 7/6 x 9/10 x 11/10 ..."
--------------------
And there may be a clue in all this for the sequence that started all this off at the top...

4 comments:

Arjen Dijksman said...

The Euler product for pi/2 is his 14th theorem of Various observations about infinite series. Happy Pi-day!

Pat's Blog said...

Arjen,
Thanks, I have been searching (without knowing exactly where to search)... Now I have to go read the first 13..(and any afters)...

Arjen Dijksman said...

Hello Pat,

I'm glad I could help in that. Your post brought me to look at Euler's works at the Euler archive, where I opened the descriptions of the works mentioning subjects related to the circle.

As an indirect consequence, it also lead me to consider infinite series approximations of pi, which in turn lead to an alternative geometrical interpretation for Wallis' product.

Pat's Blog said...

Eddie Kent wrote me that he couldn't get his comment to post, so here it is...sorry for the inconvenience... (pat)

"As far as I know he wrote it in a letter in about 1750. He also had e to 15 places - see John H. Conway & Richard K. Guy, The Book of Numbers, Springer 1996"




Eddie