README

Google Maps displays information about traffic conditions across an area. This package provides functions to produce georeferenced rasters from real-time Google Maps traffic information. Having Google traffic information in a georeferenced data format facilitates analysis of traffic information (e.g., spatially merging traffic information with other data sources).

This package was inspired by (1) existing research that has used Google traffic information, including in New York City and Dar es Salaam, and (2) similar algorithms implemented in JavaScript and in a C shell script.

Pixel values in rasters are derived from Google traffic colors and can be one of four values:

Google Traffic Color	Description	Raster Value
Green	No traffic delays	1
Orange	Medium traffic	2
Red	High traffic	3
Dark Red	Heavy traffic	4

Installation

install.packages("googletraffic")

API Key

Quickstart

Setup

## Load package
library(googletraffic)

## Load additional packages to run below examples
library(ggplot2)
library(dplyr)
library(raster)

## Set API key
google_key <- "GOOGLE-KEY-HERE"

Raster around point

To create a raster around a point, we set the centroid coordinate, the zoom level, and the height/width around the centroid coordinate (height/width are in terms of pixels, and kilometer distance of a pixel is determined primarily by the zoom level).

## Make raster
r <- gt_make_raster(location   = c(40.712778, -74.006111),
                    height     = 2000,
                    width      = 2000,
                    zoom       = 16,
                    google_key = google_key)

## Plot
r_df <- rasterToPoints(r, spatial = TRUE) %>% as.data.frame()
names(r_df) <- c("value", "x", "y")

ggplot() +
  geom_raster(data = r_df, 
  aes(x = x, y = y, 
  fill = as.factor(value))) +
  labs(fill = "Traffic\nLevel") +
  scale_fill_manual(values = c("green2", "orange", "red", "#660000")) +
  coord_quickmap() + 
  theme_void() +
  theme(plot.background = element_rect(fill = "white", color="white"))

Raster around polygon

We can also create a raster using a polygon to define the location. If needed, the function will make multiple API calls to cover the area within the polygon (a larger zoom value will result in needing to make more API calls).

## Grab shapefile of Manhattan
us_sp <- getData('GADM', country='USA', level=2)
ny_sp <- us_sp[us_sp$NAME_2 %in% "New York",]

## Make raster
r <- gt_make_raster_from_polygon(polygon    = ny_sp,
                                 zoom       = 16,
                                 google_key = google_key)

## Plot
r_df <- rasterToPoints(r, spatial = TRUE) %>% as.data.frame()
names(r_df) <- c("value", "x", "y")

ggplot() +
  geom_raster(data = r_df, 
  aes(x = x, y = y, 
  fill = as.factor(value))) +
  labs(fill = "Traffic\nLevel") +
  scale_fill_manual(values = c("green2", "orange", "red", "#660000")) +
  coord_quickmap() + 
  theme_void() +
  theme(plot.background = element_rect(fill = "white", color="white"))

Usage

See this vignette for additional information and examples illustrating how to use the package.

googletraffic