README.md

In briandconnelly/pronto: Retrieve Data from Seattle's Pronto Cycle Sharing System

pronto

pronto is a simple R package for interacting with data from from Seattle's Pronto cycle sharing system. Data comes from Pronto's data stream, which is described here.

This project is released with a Contributor Code of Conduct. By participating in this project you agree to abide by its terms.

Installation

pronto is not on CRAN, but you can install the latest and greatest version using devtools:

if(!require("devtools")) install.packages("devtools")                       
devtools::install_github("briandconnelly/pronto")

Examples

Get current station data

library(pronto)

s <- pronto_stations()

The result, s, is a list containing a timestamp for the data, whether or not rentals across the system are suspended (schemeSuspended), and a data frame containing information about all of the stations (stations).

The first five rows of stations look like this:

| id| s | n | st| b | su | m | lu| lc| bk | bl | la| lo| da| dx| ba| bx| |----:|:-----------------------|:-------|----:|:------|:------|:------|-------------:|-------------:|:------|:------|---------:|----------:|----:|----:|----:|----:| | 1| 3rd Ave & Broad St | BT-01 | 1| FALSE | FALSE | FALSE | 1.458057e+12| 1.458066e+12| FALSE | FALSE | 47.61842| -122.3510| 10| 0| 7| 1| | 2| 2nd Ave & Vine St | BT-03 | 1| FALSE | FALSE | FALSE | 1.458058e+12| 1.458066e+12| TRUE | TRUE | 47.61583| -122.3486| 11| 0| 4| 1| | 3| 6th Ave & Blanchard St | BT-04 | 1| FALSE | FALSE | FALSE | 1.458064e+12| 1.458066e+12| TRUE | TRUE | 47.61609| -122.3411| 8| 0| 7| 1| | 4| 2nd Ave & Blanchard St | BT-05 | 1| FALSE | FALSE | FALSE | 1.458059e+12| 1.458066e+12| TRUE | TRUE | 47.61311| -122.3442| 7| 0| 5| 1| | 5| 2nd Ave & Pine St | CBD-13 | 1| FALSE | FALSE | FALSE | 1.458062e+12| 1.458066e+12| TRUE | TRUE | 47.61018| -122.3396| 9| 1| 6| 2|

Getting information for a single station

Although Pronto's API doesn't support querying a single station, we can easily filter the station data using dplyr.

library(pronto)
library(dplyr)

s <- pronto_stations()

# Get information about the station near Fred Hutch
s_fhcrc <- s$stations %>%
    filter(id == 22)

Mapping available bikes

Let's make a map of current bike availability across the city using ggmap.

library(pronto)
library(dplyr)
library(magrittr)
library(ggplot2)
library(ggmap)

s <- pronto_stations()

avail <- s$stations %>%
    filter(su == FALSE) %>%
    summarise(StationsAvail=n(), BikesAvail=sum(ba))

map <- get_map(location = c(lon=mean(range(s$stations$lo)),
                            lat=mean(range(s$stations$la))),
               zoom = 13, maptype = "toner-lite")
p <- ggmap(map) +
    geom_point(data = s$stations,
               aes(x=lo, y=la, size=ba, color=ba), alpha = 0.6) +
    scale_size_area(guide=FALSE) +
    scale_color_continuous(name = "Bikes") +
    scale_alpha_continuous(guide = FALSE) +
    ggtitle(sprintf("%d bikes available at %d stations", avail$BikesAvail, avail$StationsAvail)) +
    theme_minimal() +
    theme(axis.text = element_blank()) +
    theme(axis.title = element_blank()) +
    theme(legend.title = element_text(size = rel(0.8))) +
    theme(legend.text = element_text(size = rel(0.6))) +
    theme(legend.key.size = unit(0.8, "lines"))
p

An animated version

Let's spice it up. Here, we'll get the station data every 60 seconds for one hour and then create an animated map using gganimate. Note: this code will take an hour to run.

For extra credit, we could show the time stamps in a more friendly format or interpolate the data using tweenr.

library(pronto)
library(dplyr)
library(ggplot2)
library(ggmap)
library(gganimate)

stationdata <- data.frame()
for (i in seq(60)) {
    s <- pronto_stations()
    s$stations$timestamp <- s$timestamp
    stationdata <- bind_rows(stationdata, s$stations)
    Sys.sleep(60)
}

map <- get_map(location = c(lon=mean(range(s$stations$lo)), lat=mean(range(s$stations$la))),
               zoom = 13, maptype = "toner-lite")
p <- ggmap(map) +
    geom_point(data = stationdata,
               aes(x=lo, y=la, size=ba, color=ba, frame = timestamp),
               alpha = 0.6) +
    scale_size_area(guide=FALSE) +
    scale_color_continuous(name = "Bikes") +
    scale_alpha_continuous(guide = FALSE) +
    theme_minimal() +
    theme(axis.text = element_blank()) +
    theme(axis.title = element_blank()) +
    theme(legend.title = element_text(size = rel(0.8))) +
    theme(legend.text = element_text(size = rel(0.6))) +
    theme(legend.key.size = unit(0.8, "lines"))
gg_animate(p, pause = 0.5, title_frame = FALSE)

Finding the closest bike

We've just picked up some wine at Pete's and need to get to the party. To find the nearest bike, we can use fossil to calculate the distances between our location (here) and each station. Unfortunately, this won't tell you about hills.

library(pronto)
library(fossil)
library(dplyr)
library(magrittr)

here <- list(lo = -122.329401, la = 47.639821)

closest_station <- pronto_stations()$stations %>%
    mutate(dist_km = deg.dist(.$lo, .$la, here$lo, here$la)) %>%
    arrange(dist_km) %>%
    filter(st == 1) %>%          # the station is in service
    filter(ba > 0) %>%           # there's a bike available
    head(n = 1)

cat(sprintf("The %s station is %.02f km away and has %d bike(s) available",
            closest_station$s, closest_station$dist_km, closest_station$ba))
#> The E Blaine St & Fairview Ave E station is 0.59 km away and has 5 bike(s) available

Disclaimer

Neither this package nor its contributer(s) are affiliated with Pronto.

briandconnelly/pronto documentation built on May 13, 2019, 5:14 a.m.