I want to prove "If a point lies on the interior of an angle, that
   angle is the sum of two smaller angles with legs that go through the
   given point"

Binyameen, don't know the answers to your questions, although it
sounded like John did, but let me add some mathematical comments.

1) This is a hard result in axiomatic geometry, one that Euclid could
not prove.  I discuss how Hilbert's axioms handle this in my paper
http://www.math.northwestern.edu/~richter/hilbert.pdf
Just defining the interior of an angle requires serious Hilbert theory,
and this is not done by Euclid or high school geometry texts.  So this
is a fine problem for you to be having.

2) Your ANGLES_ADD_BETWEEN come from hol_light/Multivariate/geom.ml:
John Harrison said that Multivariate is about using geometry in the
coordinate plane.  So you should first ask how you would prove this in
the plane.  Do the math first and then worry about proof assistants.

3) My hol_light/Multivariate/geom.ml, obtained by subversion, says

let ANGLES_ADD_BETWEEN = prove
 (`!A B C D:real^N.
        between C (A,B) /\ ~(D = A) /\ ~(D = B)
        ==> angle(A,D,C) + angle(C,D,B) = angle(A,D,B)`,

I don't know where between is defined, but this result makes pretty good
sense to me.  It's similar to what in my paper is called the converse of
the crossbal theorem.  Start with a triangle ADB, and suppose that C
lies in the open interval (A, B) of points between A and B.  Then C is
in the interior of angle ADB, so the measures should add up:
mu(angle ADB) = mu(angle ADC) + mu(angle CDB).  

4) You should think about how angle(A,D,B) is even defined.  It must
be something like this.  Let D = (0,0), and A = (1,0), so B = (x, y)
is essentially (cos theta, sin theta).  Then 
angle(A,D,B) = theta, right?  Let's simplify by saying that B is in
the first quadrant, with theta in (0, pi/2).  
Well, tan theta = y/x, so 
theta = tan^{-1} (y/x) 
So John's definition is quite possibly 

angle((1,0),(0,0),(x,y)) = tan^{-1} (y/x) 

for x, y > 0.  If you're willing to talk about directed angles, as
Birkhoff did, then you can use this formula for x > 0, and the angle
will be negative for y < 0.

-- 
Best,
Bill 

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
hol-info mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/hol-info

Reply via email to