Skip to content

I'm making this repo for geohashing enthusiasts who want to encode and decode geohashes in the most python-optimized way. It's now part of the Pygeohash package yay!

Notifications You must be signed in to change notification settings

IlyasMoutawwakil/geohash-on-steroids

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

13 Commits
 
 
 
 
 
 
 
 

Repository files navigation

geohash-on-steroids

I'm making this repository for geohashing enthusiasts who want to encode and decode geohashes in the most python-optimized way. I'll be experimenting more and adding more fuctionalities very soon.

Dependencies

Optimized functions are created with the njit decorators and using arrays so the only dependencies are Numba and Numpy. You can install them using pip in any python environment: pip install numpy numba

Performance

As you can see in my notebook, performance gain in comparison to what's on the python package pygeohash is the following:

# PyGeoHash function
%%timeit
point_decode(geohash)
# Output: 20.4 µs ± 367 ns per loop (mean ± std. dev. of 7 runs, 10000 loops each)
# Numba function
%%timeit
nb_point_decode(geohash)
# Output: 4.48 µs ± 16.8 ns per loop (mean ± std. dev. of 7 runs, 100000 loops each)
# PyGeoHash function
%%timeit
point_encode(latitude, longitude)
# Output: 92.8 µs ± 2.37 µs per loop (mean ± std. dev. of 7 runs, 10000 loops each)
# Numba function
%%timeit
nb_point_encode(latitude, longitude)
# Output: 11.2 µs ± 663 ns per loop (mean ± std. dev. of 7 runs, 100000 loops each)

But geohashing is generally performaded on large amounts of data points so I made a vector-wise implimentation that performs even better at larger scale:

# Numpy vectorization function
%%timeit
np_vector_decode(geohashes)
# Output: 2.09 s ± 25.6 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)
# Numba vectorization function
%%timeit
nb_vector_decode(geohashes)
# Output: 164 ms ± 1.66 ms per loop (mean ± std. dev. of 7 runs, 10 loops each)
# Numpy vectorization function
%%timeit
np_vector_encode(latitudes, longitudes)
# Output: 2.57 s ± 53.8 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)
# Numba vectorization function
%%timeit
nb_vector_encode(latitudes, longitudes)
# Output: 443 ms ± 12.7 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)

Roadmap

  • Use Numba for computing efficient geohash encoding and decoding
  • Use Numba for computing efficient geohash decoding
  • Use Numba for vector-wize computing efficient geohash encoding
  • Use Numba for vector-wize computing efficient geohash decoding
  • Parallelize computations efficiently (one loop? no loops?)

About

I'm making this repo for geohashing enthusiasts who want to encode and decode geohashes in the most python-optimized way. It's now part of the Pygeohash package yay!

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published