tiling.txt

Tiling process

/base_url/zoom/x/y.json

zoom 0 is full world
runs from -180 to 180 longitude
runs from -90 to 270 latitude

size of tile is 360/2**zoom level

west end of tile = -180+(x*(size)/2)
east end of tile = west end + size
south = -90+(y*(size)/2)
north = south + size

three ways to represent data:
      1. raw data store
      2. pre-tile representation (optional, goal)
      3. actual json tile

cells per tile, must be even, 10x10, should not be hardcoded

overall code structure:

1. database implementation for data
   1a. generate random test data
   1b. use sqlite for now

2. generating tiles
   for pt in pts:
       for tiles containing pt:
       	   write pt to file
   write out all tiles

calculating which tiles a point is in:

tile.west < point.lng < tile.east
tile.south < point.lat < tile.north

-180+(x*(360/(2**(z+1)))  = point.lng

(x*(360/(2**(z+1))) = point.lng+180

x = (point.lng+180)/(360/(2**(z+1)))
y = (point.lat+ 90)/(360/(2**(z+1)))

take floor x, you just got tile index!

(using current settings for defaultSettings.py)

for zoom in range(10):
    xf = (point.lng+180)/(360/(2**(z+1)))
    yf = (point.lat+ 90)/(360/(2**(z+1)))
    xmax = math.floor(xf)
    ymax = math.floor(yf)
    xmin = max(0,xmax-1)
    ymin = max(0,ymax-1)
    xcell = math.floor(((xf-xmax)*CELLS_PER_TILE)*2)
    ycell = math.floor(((yf-ymax)*CELLS_PER_TILE)*2)
    for x in range(xmin,xmax+1):
    	for y in range(ymin,ymax+1):
	    tile = tiles[zoom,x,y]
	    tile[xcell,ycell].process(point)