Geohashing functions

The spatio-temporal library includes geohashing functions for proximity search (encoding latitude and longitude and grouping nearby points) in location data analysis.

Geohash encoding

  • Encode a Point object with a geographic location to a base-32 encoded string of letters or digits:

    >>> p = stc.point(37, -74)
>>> stc.geohash.string_hash_encode(p)
'dqe6kpdue5btnubpm9npcd0000'
>>> stc.geohash.string_hash_encode(p, precision=5)
'dqe6k'

  • Encode a Point object with a geographic location to a BitVector:

    >>> p = stc.point(37, -74)
>>> bv = stc.geohash.number_hash_encode(p)
>>> bv
BitVector(01100101100110100110100101010101100110100110100101010101100110100110100101010101100110100110100101010101101100000000000000000000)
>>> bv.truncate(10)
>>> bv
BitVector(0110010110)

  • Encode a pair of latitude and longitude coordinates directly to a geohash:

    >>> stc.geohash.encode(37, -74, precision=5, mode='base32')
'dqe6k'
>>> stc.geohash.encode(37, -74, precision=10, mode='bit')
'0110010110'

Geohash decoding

  • Decode geographic base 32 digits or a string to a Point object with the decimal longitude or latitude coordinates:

    >>> stc.geohash.string_hash_decode('dqeh4')
Point(37.265625, -74.443359375)

  • Decode base-32 encoded characters to a Point object:

    >>> p = stc.point(37, -74)
>>> bv = stc.geohash.number_hash_encode(p)
>>> stc.geohash.number_hash_decode(bv)
Point(37.0, -74.0)

  • Decode a base-32 or binary string to a pair of latitude and longitude coordinates:

    >>> stc.geohash.decode('dqe6k', mode='base32')
(36.9580078125, -74.00390625)
>>> stc.geohash.decode('0110010110', mode='bit')
(33.75, -78.75)

Geohash neighbors

The geohash neighbors function returns the neighboring geohash codes around a specified code.

Geohash neighbors for a point

  • Get the geohash neighbors for the given latitude, longitude and bit_depth:

    >>> stc.geohash.get_all_neighbors(70, -40, 25)
[BitVector(0111110000001111100101010),
BitVector(0111110000001111100101011),
BitVector(0111110000001111100101110),
BitVector(0111110000001111110000000),
BitVector(0111110000001111110000001),
BitVector(0111110000001111110000100),
BitVector(0111110000001111110000010),
BitVector(0111110000001111110000011),
BitVector(0111110000001111110000110)]

  • Get the geohash neighbors for a given latitude, longitude and bit_depth, and only return the results that are within a given distance:

    >>>stc.geohash.get_all_neighbors(70, -40, 25, distance_precision=1000)
[BitVector(0111110000001111110000001),
BitVector(0111110000001111110000100),
BitVector(0111110000001111110000011),
BitVector(0111110000001111110000110)]

Geohash neighboring functions

  • expand: Returns the center code and its neighbors. All 9 geohashes for a location are returned:

    >>> stc.geohash.expand('ezs42')
['ezefp', 'ezs40', 'ezs41', 'ezefr', 'ezs42', 'ezs43', 'ezefx', 'ezs48', 'ezs49']

  • neighbors: Returns 8 geohashes, excluding the given geohash itself:

    >>> stc.geohash.neighbors('ezs42')
['ezefp', 'ezs40', 'ezs41', 'ezefr', 'ezs43', 'ezefx', 'ezs48', 'ezs49']

  • get_east, get_west, get_north, get_south: Returns the east, west, north, or south geohash for a given geohash:

    >>> stc.geohash.get_east('ezs42')
'ezs43'

  • Encode a geohash to a BitVector:

    >>> bv = stc.geohash.number_hash_encode(p)
>>> bv.truncate(25)
>>> stc.geohash.expand(bv)
[BitVector(0110010110011010011000101),
BitVector(0110010110011010011010000),
BitVector(0110010110011010011010001),
BitVector(0110010110011010011000111),
BitVector(0110010110011010011010010),
BitVector(0110010110011010011010011),
BitVector(0110010110011010011001101),
BitVector(0110010110011010011011000),
BitVector(0110010110011010011011001)]
>>> stc.geohash.neighbors(bv)
[BitVector(0110010110011010011000101),
BitVector(0110010110011010011010000),
BitVector(0110010110011010011010001),
BitVector(0110010110011010011000111),
BitVector(0110010110011010011010011),
BitVector(0110010110011010011001101),
BitVector(0110010110011010011011000),
BitVector(0110010110011010011011001)]
>>> stc.geohash.get_north(bv)
BitVector(0110010110011010011011000)

Geohash coverage

To calculate a set of geohashes that wholly covers the bounding box:

  1. Prepare a sample polygon:

    test_wkt = 'POLYGON((-73.76223024988917 41.04173285255264,-73.7749331917837 41.04121496082817,-73.78197130823878 41.02748934524744,-73.76476225519923 41.023733725449326,-73.75218805933741 41.031633228865495,-73.7558787789419 41.03752486433286,-73.76223024988917 41.04173285255264))'
poly = wkt_reader.read(test_wkt)

  2. Compute the geohash cover at a fixed bit depth for a given geometry:

    cover = stc.geohash.geohash_cover_at_bit_depth(poly, 36)

    Shows geohash coverage at a fixed bit depth for a given geometry.

  3. Compute the buffered geohash cover at a fixed bit depth for a given geometry:

    buffered_cover = stc.geohash.geohash_cover_at_bit_depth(poly, 36, distance=50)

    Shows buffered geohash coverage at a fixed bit depth for a given geometry.

  4. Compute a compact geohash cover by computing the fixed depth cover first and then compressing the cover:

    raw_cover = stc.geohash.geohash_cover_at_bit_depth(poly, 36)
compact_cover = stc.geohash.geohash_compression(raw_cover)

    Shows compact geohash coverage at a fixed bit depth for a given geometry.

Parent topic: Geospatio-temporal library