OpenEuphoria

Re: Best way to...

• Posted by ArthurCrump Oct 19, 2010
• 1690 views

I should get out of the habit of doing things in too much of a hurry.
In my previous reply, I forgot that sin and cos want radians, not degrees.
Here is a function that takes the list of airports and a current location as degrees East and North. It seemed to get the distance from San Francisco to Los Angeles airport about right. It uses Euphoria version 4.

include std/sequence.e 
include std/math.e 
include std/convert.e 
constant EarthRadius = 3963		-- Approximate miles, or 6378 Kilometres 
-- Parameters of 'nearest': 
-- site_list: The list of airports, in the original form, {"KLAX|-118.4etc|33.94etc",...} 
-- E: The Easting of the current location, ie degrees longitude East. 
-- N: The Northing of the current location, ie degrees latitude North. 
public function nearest(sequence site_list, atom E, atom N ) 
	sequence bestsite, this, XYZhere, XYZsite 
	atom bestcp2, ChordSquared 
	integer bestindex 
	N = deg2rad(N) 
	E = deg2rad(E) 
	XYZhere = { cos(N)*cos(E), cos(N)*sin(E), sin(N) } 
	bestcp2 = 9.9		-- The actual range is 0.0 to 2.0 
	for n = 1 to length(site_list) do 
		-- First convert "name|degrees|degrees" to {"name",radians, radians} 
		this = split(site_list[n],'|') 
		this[2] = deg2rad(to_number(this[2])) 
		this[3] = deg2rad(to_number(this[3])) 
		-- It would be more efficient if site_list entries were already in this form. 
		XYZsite = { cos(this[3])*cos(this[2]), cos(this[3])*sin(this[2]), sin(this[3]) } 
		ChordSquared = sum(power(XYZsite-XYZhere,2)) 
		if ChordSquared < bestcp2 then 
			bestcp2 = ChordSquared 
			bestsite = this 
		end if 
	end for 
	return bestsite & 2*arcsin(sqrt(bestcp2)/2)*EarthRadius & sqrt(bestcp2) 
end function 
-- The function returns a sequence of 4 elements for the nearest airport: 
-- {Name, Easting, Northing, Surface_distance}		-- Currently in miles. 
-- To take account of polar flattening, the Z coordinate should be multiplied by 
-- 0.99665. The chord lengths and the nearest airport would be correct, but the 
-- surface distance would be more difficult to work out from the chord length. 
 

Arthur