Showing posts with label unit circle. Show all posts
Showing posts with label unit circle. Show all posts

Wednesday, 17 December 2008

Another Approach to Pythagorean Triples

Click on image to enlarge

I wrote a couple of posts a while back on the Barning Tree method of finding Pythagorean triples using matrices, and then a followup. I recently came across another approach to Pythagorean triples that involves a clever relationship between points on the positive y-axis, and points on the unit circle. (ok, maybe I should have known this, but I didn't.. and I think it is really a neat idea)

On the unit circle, x2 + y 2 = 1, if we draw a secant from the point (-1,0) through (0,t) on the y-axis, it turns out that if t is a rational number, then the coordinates of P=(x,y) where the secant intersects the circle, will also be rational. Since the slope of the line is also t, the equation is y=tx+t ... and so and t2= (1-x2)/(1+x)2. That means t = y/(x+1) which leads to x= (1-t^2)/(1+t^2) and y= (2t)/(1+t^2) If we pick some rational number to be t, say t=2/7,
then x= 45/53 and y= 28/53.... Then by similar triangles, there must be a circle with radius 53 and a point on the circle would be (28, 45) and in fact 282+452=532... and any such rational point will produce another Image of unit circle