GoBike Data

This project visualizes data about the Ford GoBike network. The data is available here: https://bikeshare.science.

Trip Data

Trip data is downloaded from https://www.fordgobike.com/system-data and should be placed in the data directory. Once downloaded, the directory should look like this:

$ ll data
total 1309768
drwxr-xr-x  11 kevin  staff   352B Aug 14 01:08 .
drwxr-xr-x  22 kevin  staff   704B Aug 19 20:09 ..
-rw-r--r--@  1 kevin  staff   112M Aug 12 02:44 2017-fordgobike-tripdata.csv
-rw-r--r--@  1 kevin  staff    19M Aug 12 02:44 201801-fordgobike-tripdata.csv
-rw-r--r--@  1 kevin  staff    22M Aug 12 02:44 201802-fordgobike-tripdata.csv
-rw-r--r--@  1 kevin  staff    23M Aug 12 02:44 201803-fordgobike-tripdata.csv
-rw-r--r--@  1 kevin  staff    27M Aug 12 02:44 201804-fordgobike-tripdata.csv
-rw-r--r--@  1 kevin  staff    36M Jun  8 08:08 201805-fordgobike-tripdata.csv
-rw-r--r--@  1 kevin  staff    40M Jul 16 11:40 201806-fordgobike-tripdata.csv
-rw-r--r--@  1 kevin  staff    40M Aug  7 12:01 201807-fordgobike-tripdata.csv

This is a prerequisite for building the site.

Static Site

All of the pages are static pages that are checked in to Git. Run make site to regenerate the HTML pages.

Testing

Run make test to run the test suite.

Polygons

The polygons are kind of a pain. Use geojsonlint to check whether your polygons are okay. They need to be in a particular order.

Run the rewind script to rewind the polygon order.

Datasets

In addition to server, we've made the Ford GoBike datasets available online.

BigQuery

All trip data lives in the ford_gobike dataset, which is available publicly.

Trips per week

SELECT
  DATE_TRUNC(DATE(start_time), WEEK) as week,
  COUNT(*) as trips
FROM `bay-area-public-data.ford_gobike.trips`
GROUP BY 1
ORDER BY 1

Unique bikes per week

SELECT
  DATE_TRUNC(DATE(start_time), WEEK) as week,
  COUNT(distinct bike_id) as bikes
FROM `bay-area-public-data.ford_gobike.trips`
GROUP BY 1
ORDER BY 1

Average trips per bike per week

WITH bike_trips AS (
SELECT
  DATE_TRUNC(DATE(start_time), WEEK) as week,
  bike_id,
  count(*) as trips
FROM `bay-area-public-data.ford_gobike.trips`
GROUP BY 1, 2
ORDER BY 1
)

SELECT week, avg(trips)
FROM bike_trips
GROUP BY 1

kevinburke / gobike

readme