Difference between revisions of "Known Issues"

From Geohashing
imported>Robyn
(Categories are cool.)
imported>Zamzara
(Dow values shorter than five digits)
 
(5 intermediate revisions by 5 users not shown)
Line 1: Line 1:
 
'''Please direct all discussion on these issues to [[Talk:Main Page]].'''
 
'''Please direct all discussion on these issues to [[Talk:Main Page]].'''
  
*Due to the relative start locations used, if you happen to live close to a coastal region you may require a boat in order to successfully reach the calculated coordinates.   
+
* Due to the relative start locations used, if you happen to live close to a coastal region you may require a boat in order to successfully reach the calculated coordinates.   
  
*In addition, due to the relative start locations used, you may require a passport to successfully reach the calculated coordinates.
+
* In addition, due to the relative start locations used, you may require a passport to successfully reach the calculated coordinates.
  
 
* Some of the mapping URLs don't cope with -0 as entry, making this grid (eg London West) difficult to map. See [[-0 Issue]].
 
* Some of the mapping URLs don't cope with -0 as entry, making this grid (eg London West) difficult to map. See [[-0 Issue]].
Line 11: Line 11:
 
* <s>The algorithm tends to produce points clustered near the graticule origin, though not on it. For a more random result, try ignoring the first digit after converting to decimal (try plotting the points from a whole year's worth of data to see this)</s> see talk page
 
* <s>The algorithm tends to produce points clustered near the graticule origin, though not on it. For a more random result, try ignoring the first digit after converting to decimal (try plotting the points from a whole year's worth of data to see this)</s> see talk page
  
* There is a small bias towards the pole side of the graticule. The graticules are 1 degree longitude by 1 degree lattitude, but 1 degree longitude is smaller near the poles than near the equator. Since the [[algorithm]] does not use/know this information, a point at the pole side of the graticule has more chance of becoming a geogash location than a point at the equator side. In other words and more general: a square (kilo)meter near the pole has more chance of 'hosting' a geohash location than a square (kilo)meter near the equator.
+
* There is a bias in hash density towards the pole side of the graticule. The graticules are 1 degree longitude by 1 degree lattitude, but 1 degree longitude is smaller near the poles than near the equator. Since the [[algorithm]] does not account for this information, a given area at the pole side of the graticule has a greater chance of containing a geohash location than the same size area at the equator side. Similarly, a given area in a graticule closer to a pole has a greater chance of containing a geohash than the same area in a graticule closer to the equator.
[[Category:The Algorithm]]
+
 
 +
* For locations west of [[30W]], on Friday the points for the Weekend are known. For locations east of [[30W]], this only happens during Friday and can be on Saturday, because only on their Saturday the new [[Dow Jones Industrial Average]] opening from Friday will become available. So, the people from the west have an advantage of having one day more for planning meetups than people from the east, not mentioning the cases for people out of internet during the weekend.
 +
 
 +
* When calculating some [[retro]] dates, the [[Dow]] value will be fewer than five digits long. These are not padded with a leading zero — e.g. 1234.56, not 01234.56.
 +
 
 +
[[Category:Algorithm]]

Latest revision as of 11:57, 17 August 2015

Please direct all discussion on these issues to Talk:Main Page.

  • Due to the relative start locations used, if you happen to live close to a coastal region you may require a boat in order to successfully reach the calculated coordinates.
  • In addition, due to the relative start locations used, you may require a passport to successfully reach the calculated coordinates.
  • Some of the mapping URLs don't cope with -0 as entry, making this grid (eg London West) difficult to map. See -0 Issue.
  • When using Firefox w/ Firebug there is an error that shows up when clicking on a marker on the map. It seems the handler is trying to read a property called 'latlng' which isnt defined when clicking on a marker. Line 81 on xkcd.js. Maybe if latlng == undefined, then either exit ( if nothings supposed to happen ) or use the getLatLng() function on the passed object.
  • The algorithm tends to produce points clustered near the graticule origin, though not on it. For a more random result, try ignoring the first digit after converting to decimal (try plotting the points from a whole year's worth of data to see this) see talk page
  • There is a bias in hash density towards the pole side of the graticule. The graticules are 1 degree longitude by 1 degree lattitude, but 1 degree longitude is smaller near the poles than near the equator. Since the algorithm does not account for this information, a given area at the pole side of the graticule has a greater chance of containing a geohash location than the same size area at the equator side. Similarly, a given area in a graticule closer to a pole has a greater chance of containing a geohash than the same area in a graticule closer to the equator.
  • For locations west of 30W, on Friday the points for the Weekend are known. For locations east of 30W, this only happens during Friday and can be on Saturday, because only on their Saturday the new Dow Jones Industrial Average opening from Friday will become available. So, the people from the west have an advantage of having one day more for planning meetups than people from the east, not mentioning the cases for people out of internet during the weekend.
  • When calculating some retro dates, the Dow value will be fewer than five digits long. These are not padded with a leading zero — e.g. 1234.56, not 01234.56.