geohash-iOS is an implementation of GeoHash based spacial proximity calculation in Objective-C to be used for mobile applications. It contains the necessary classes and a tester to check the functionality.

A GeoHash is a representation for a certain location on the world map. It is encoded through an amount of base32 characters depending on a given precision. The string can be transfered into a binary representation where the even bits are extracted to calculate longitude and the odd bits to calculate latitude correspondingly.

GeoHashes are easy to handle compared to the numerical latitude/longitude values. In that way they are good for data processing and storage of point data.

A good detailled description can be found at Wikipedia.

To check the position of a hash, go to http://geohash.org. Add the hash string to the URL (e.g. for hash r3gqm68z9gnw enter http://geohash.org/r3gqm68z9gnw as URL.

The precision of a GeoHash increases with each additional character. Assuming the first character defines a large area on the map of e.g. 1000 sqkm, an additional second character will define a smaller area of e.g. 100 sqkm within the area defined by the first character. In that way the chain of characters can be regarded as hierachical nested area definitions with each additional character defining a point/area more precise within a larger area. A good visual description of the concept is given in this presentation. An interactive map with areas visualized depending on hash precision can be found at openlocation.org.

According to the hierarchical definition of subareas in a hash string it is easy to identify whether different hashes are within a certain proximity. E.g. given hashes r3gqm68z9gnw and r3gqm68z9gnx and a required precision of e.g. 12 it can be assumed the hashes to be quite close to each other since only the last character differs while a hash u33dc0f42e445 would far away.

At GNS Science the idea is used to identify whether the location of an appearing or expected earthquake is close to any cities. An iPhone App can be used to find occurences close to certain areas.

The main-method is to be found in the supporting files folder in the file main.m. The test method is testGeoHashes in the GNViewController class.

The GitHub page now has the 3 main classes for everyone interested in using GeoHashes in their iOS project to use right on the main page. In the 'demo' subdirectory is a complete XCode project that uses the 3 main classes and implements a little demo - on the iPhone or iPod Touch a map is shown with a Pin on it - the Pin shows the location on the map (lat,lon), as well as the corresponding GeoHash. We calculate the GeoHash by default to 12 characters and display the area that corresponds to 2 characters accuracy (otherwise it's too small) as a shaded rectangle under the pin.

float lat = -34.042580;
float lon = 151.052139;
int precision = 12;

GNGeoHash *gh = [GNGeoHash withCharacterPrecision:lat andLongitude:lon andNumberOfCharacters:precision];

//print the result
NSLog(@"GeoHash is %@",[gh toBase32]);

Result is

GeoHash is r3gqm68z9gnw

GNGeoHash *gh = [GNGeoHash withCharacterPrecision:lat andLongitude:lon andNumberOfCharacters:precision];

GNGeoHash *northern = [gh getNorthernNeighbour];
GNGeoHash *eastern = [gh getEasternNeighbour];
GNGeoHash *southern = [gh getSouthernNeighbour];
GNGeoHash *western = [gh getWesternNeighbour];

//print results
NSLog(@"Northern hash is %@",[northern toBase32]);
NSLog(@"Eastern hash is %@",[eastern toBase32]);
NSLog(@"Southern hash is %@",[southern toBase32]);
NSLog(@"Western hash is %@",[western toBase32]);

Result is

Northern hash is r3gqm68z9gnx
Eastern hash is  r3gqm68z9gny
Southern hash is r3gqm68z9gnt
Western hash is W has GeoHash r3gqm68z9gnq

GNGeoHash *gh = [GNGeoHash withCharacterPrecision:lat andLongitude:lon andNumberOfCharacters:precision];

NSLog(@"Upper left corner is %@",[gh.boundingBox getUpperLeft]);
NSLog(@"Lower right corner is %@",[gh.boundingBox getLowerRight]);

Result is

Upper left corner is -34.042580,151.052139
Lower right corner is -34.042580,151.052140

• Latitude values have to be between -90 and 90
• Longitude values have to be between -180 and 180
• Precision has to be between 1 and 12

• iOS 5.1

(This project is released under the GPL)

Copyright (c) 2012 GNS Science. All rights reserved. http://www.gns.cri.nz/