Linked Plot Maps for Multivariate High Resolution Spatio-Temporal Data • DrivePlotR

The goal of DrivePlotR is to produce linked plot maps for multivariate high resolution spatio-temporal data such as data from naturalistic driving studies and connected vehicles.

Installation

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

# install.packages("pak")
pak::pak("hardtme/DrivePlotR")

Example

This example shows the basic workflow for making visualizations with DrivePlotR.

Load DrivePlotR and make drive data available in the R session. We have made non-sensitive drives from a naturalistic driving study available as part of the DrivePlotR package, but any other data with both geographic and temporal data would also be suitable.

library(DrivePlotR)

data("nds_data")
drive <- nds_data |> dplyr::filter(drive == 7)

Establish the geographic variables and projection. The code below creates a simple feature variable geometry of all points. Setting the geographic projection to WGS84 ensures that points in the data line up with maps in Leaflet.

drive <- drive |> 
  sf::st_as_sf(coords = c("gps_long", "gps_lat"), 
               crs = "WGS84")

Convert the drive data into a shared data frame.

shared_drive <- 
  crosstalk::SharedData$new(drive)

Create visualizations.

Although the main purpose of DrivePlotR is for users to create linked visualizations with a map and one to four companion graphs, users can also separately create maps or companion graphs.

You can create a basic standalone map as follows:

driveplot_map(shareddata = shared_drive)

You can also create a standalone companion graph as follows:

driveplot_companions(
  shareddata = shared_drive, 
  time = time_cst, y1 = speed_mph, 
  timelabel = "Time", 
  y1label = "Speed (MPH)")

DrivePlotR offers many options for customization. For example, the following code creates a more complex linked plot map. First, we create a new variable to track each minute of the drive. We will use this variable to color the points to help us distinguish what happened during each minute of the drive.

drive <- drive |>
  dplyr::mutate(time_cst = lubridate::ymd_hms(time_cst, tz = "US/Central"),
                gps_minute = as.factor(lubridate::minute(time_cst)))

shared_drive <- shared_drive <- 
  crosstalk::SharedData$new(drive)

The companion graphs include speed, gyroscopic heading (the direction the vehicle is going), and GPS position dilution of precision (PDOP, which is a measure of the quality of the GPS location data). The points on both the map and the companion graphs are colored by the minute of the drive during which they occurred.

driveplot(
  shareddata = shared_drive, 
  maplabel = time_cst, colorvar = gps_minute, 
  colorpalette = "viridis", fillOpacity = 1, 
  time = time_cst, y1 = speed_mph, 
  y2 = gyro_heading, y3 = gps_pdop, 
  timelabel = "Time", y1label = "Speed (MPH)", 
  y2label = "Gyro Heading", 
  y3label = "GPS PDOP", showlegend = TRUE, 
  legendtitle = "Minute", 
  plottitle = "A Drive in Omaha, NE", 
  spacing = 0.05)