Polars-Reverse-Geocode

Polars plugin based on https://github.com/gx0r/rrgeo.

rrgeo takes a latitude and longitude as input and returns the closest city, country, latitude, and longitude, using a k-d tree to efficiently find the nearest neighbour based on a known list of locations. This can be useful if you need to reverse geocode a large number of coordinates quickly, or just need the rough location of coordinates but don't want the expense or complication of an online reverse geocoder.

Installation

pip install polars-reverse-geocode

Usage example

import polars as pl

from polars_reverse_geocode import find_closest_city, find_closest_state, find_closest_country

df = pl.DataFrame({
    'lat': [37.7749, 51.01, 52.5],
    'lon': [-122.4194, -3.9, -.91]
})

print(
    df.with_columns(
        city = find_closest_city('lat', 'lon'),
        state = find_closest_state('lat', 'lon'),
        country_code = find_closest_country('lat', 'lon')
    )
)

shape: (3, 5)
┌─────────┬───────────┬───────────────────┬────────────┬──────────────┐
│ lat     ┆ lon       ┆ city              ┆ state      ┆ country_code │
│ ---     ┆ ---       ┆ ---               ┆ ---        ┆ ---          │
│ f64     ┆ f64       ┆ str               ┆ str        ┆ str          │
╞═════════╪═══════════╪═══════════════════╪════════════╪══════════════╡
│ 37.7749 ┆ -122.4194 ┆ San Francisco     ┆ California ┆ US           │
│ 51.01   ┆ -3.9      ┆ South Molton      ┆ England    ┆ GB           │
│ 52.5    ┆ -0.91     ┆ Market Harborough ┆ England    ┆ GB           │
└─────────┴───────────┴───────────────────┴────────────┴──────────────┘

Caching

Loading the geocoding data set has a fixed cost of roughly 150ms, which is noticeable when doing repeated, one-off lookups (e.g. sequential/chained calls in a UI). Each find_closest_* function accepts a cache_mode keyword argument to control this caching behaviour:

• "cache_forever" (default): loads the data set once per process and keeps it in memory (~2.5GB RAM) for the lifetime of the process, so subsequent calls are fast.
• "do_not_cache": does not populate the shared, process-wide cache. If the cache has already been populated (e.g. by an earlier "cache_forever" call), it's reused; otherwise, the data set is loaded fresh on every call, and discarded afterwards, avoiding the memory cost.

df.with_columns(
    city=find_closest_city("lat", "lon", cache_mode="do_not_cache"),
)