rreversegeocoder

The goal of rreversegeocoder is to, given latitude and longitude, find the nearest city in US.

Why? We get new features for feature engineering that can improve the performance of machine learning models.

Installation

You can install the development version of rreversegeocoder from GitHub with:

# pak::pak("jrosell/rreversegeocoder")
install.packages("rreversegeocoder", repos = c('https://jrosell.r-universe.dev', 'https://cloud.r-project.org'))

Example

This is a basic example which shows you how to get all the geoname information from a given latitude and longitude in US:

library(tidyverse)
library(tidymodels)
library(rreversegeocoder)

cites <- download_cities_with_population_5000()

input <- tribble(
    ~latitude, ~longitude,
    34.6, -120,
    34.1, -100
)

output <- input |>
    bind_cols(reverse_geocoder_us_cities_with_population_5000(input)) |>
    left_join(cites, by = "geonameid")

glimpse(output)

Rows: 2
Columns: 21
$ latitude           34.6, 34.1
$ longitude          -120, -100
$ geonameid          "5397059", "4739078"
$ name               "Solvang", "Vernon"
$ asciiname          "Solvang", "Vernon"
$ alternatenames     "Solvang,Solvang i California,Solvanq,Solveng,solbhya…
$ geoname_latitude   34.59582, 34.15536
$ geoname_longitude  -120.13765, -99.26628
$ feature_class      "P", "P"
$ feature            "PPL", "PPLA2"
$ country            "US", "US"
$ cc2                NA, NA
$ admin1             "CA", "TX"
$ admin2             "083", "487"
$ admin3             NA, NA
$ admin4             NA, NA
$ population         5741, 10573
$ elevation          154, 361
$ dem                152, 361
$ timezone           "America/Los_Angeles", "America/Chicago"
$ mdate              2017-03-09, 2017-03-09