Analyzing NYC Transit Data
2 likes10,952 views
The document discusses the analysis of NYC transit data from taxis, Ubers, and Citi Bikes, focusing on data collection and processing methods, including the use of tools like PostgreSQL and R. It outlines the challenges encountered with raw data and demonstrates how to map geographical coordinates to NYC census tracts using PostGIS. Additionally, it provides insights into pre-aggregating data for efficient analysis and emphasizes the importance of engaging writing in presenting findings.
Analyzing NYC Transit Data
- 1. Analyzing NYC Transit Data: Taxis, Ubers, and Citi Bikes Todd Schneider April 8, 2016 todd@toddwschneider.com
- 2. Where to find me toddwschneider.com github.com/toddwschneider @todd_schneider toddsnyder
- 3. Things I’ll talk about • Taxi, Uber, and Citi Bike data • Medium data analysis tools and tips • Where does R fit in?
- 4. Taxi and Uber Data http://toddwschneider.com/posts/analyzing-1-1-billion-nyc-taxi-and-uber-trips-with-a-vengeance/
- 6. NYC Taxi and Uber Data • Taxi & Limousine Commission released public, trip-level data for over 1.1 billion taxi rides 2009–2015 • Some public Uber data available as well, thanks to a FOIL request by FiveThirtyEight http://www.nyc.gov/html/tlc/html/about/trip_record_data.shtml
- 7. Citi Bike Data • Citi Bike releases monthly data for every individual ride • Data includes timestamps and locations, plus rider’s subscriber status, gender, and age https://www.citibikenyc.com/system-data
- 8. Generic Analysis Overview 1. Get raw data 2. Write code to process raw data into something more useful 3. Analyze data 4. Write about what you found out
- 9. Analysis Tools • PostgreSQL • PostGIS • R • Command line • JavaScript https://github.com/toddwschneider/nyc-taxi-data https://github.com/toddwschneider/nyc-citibike-data
- 10. Raw data processing goals • Load flat files of varying file formats into a unified, persistent PostgreSQL database that we can use to answer questions about the data • Do some one-time calculations to augment the raw data • We want to answer neighborhood-based questions, so we’ll map latitude/longitude coordinates to NYC census tracts
- 11. Processing raw data: The reality • Often messy, raw data can require massaging • Not fun, takes a while, but is essential • Specifically: we have to plan ahead a bit, anticipate usage patterns, questions we’re going to ask, then decide on schema
- 12. Raw Data
- 13. Specific issues encountered with raw taxi data • Some files contain empty lines and unquoted carriage returns 😐 • Raw data files have different formats even within the same cab type 😕 • Some files contain extra columns in every row 😠 • Some files contain extra columns in only some rows 😡
- 14. How do we load a bunch of files into a database? • One at a time! • Bash script loops through each raw data file, for each file it executes code to process data and insert records into a database table https://github.com/toddwschneider/nyc-taxi-data/blob/master/import_trip_data.sh
- 15. How do we map latitude and longitude to census tracts? • PostGIS! • Geographic information system (GIS) for PostgreSQL • Can do calculations of the form, “is a point inside a polygon?” • Every pickup/drop off is a point, NYC’s census tracts are polygons
- 16. NYC Census Tracts • 2,166 tracts • 196 neighborhood tabulation areas (NTAs)
- 17. Shapefiles • Shapefile format describes geometries like points, lines, polygons • Many shapefiles publicly available, e.g. NYC provides a shapefile that contains definitions for all census tracts and NTAs • PostGIS includes functionality to import shapefiles
- 18. Shapefile Example
- 19. PostGIS: ST_Within() • ST_Within(geom A, geom B) function returns true if and only if A is entirely within B • A = pickup or drop off point • B = NYC census tract polygon
- 20. Spatial Indexes • Problem: determining whether a point is inside an arbitrary polygon is computationally intensive and slow • PostGIS spatial indexes to the rescue!
- 21. Spatial indexes in a nutshell bounding box Bounding box Census tract
- 22. Spatial Indexes • Determining whether a point is inside a rectangle is easy! • Spatial indexes store rectangular bounding boxes for polygons, then when determining if a point is inside a polygon, calculate in 2 steps: 1. Is the point inside the polygon’s bounding box? 2. If so, is the point inside the polygon itself? • Most of the time the cheap first check will be false, then we can skip the expensive second step
- 23. Putting it all together • Download NYC census tracts shapefile, import into database, create spatial index • Download raw taxi/Uber/Citi Bike data files and loop through them, one file at a time • For each file: fix data issues, load into database, calculate census tracts with ST_Within() • Wait 3 days and voila!
- 24. Analysis, a.k.a.“the fun part” • Ask fun and interesting questions • Try to answer them • Rinse and repeat
- 25. Taxi maps • Question: what does a map of every taxi pickup and drop off look like? • Each trip has a pickup and drop off location, plot a bunch of dots at those locations • Made entirely in R using ggplot2
- 26. Taxi maps
- 27. Taxi maps preprocess • Problem: R can’t fit 1.1 billion rows • Solution: preprocess data by rounding lat/long to 4 decimal places (~10 meters), count number of trips at each aggregated point https://github.com/toddwschneider/nyc-taxi-data/blob/master/analysis/prepare_analysis.sql#L194-L215
- 28. Render maps in R https://github.com/toddwschneider/nyc-taxi-data/blob/master/analysis/analysis.R
- 29. Data reliability Every other comment on reddit:
- 30. • Map the position of every Citi Bike over the course of a single day • Google Maps Directions API for cycling directions • Leaflet.js for mapping • Torque.js by CartoDB for animation Citi Bike Animation
- 31. • Google Maps cycling directions have strong bias for dedicated bike lanes on 1st, 2nd, 8th, and 9th avenues • Not necessarily true! Citi Bike Assumptions
- 32. Modeling the relationship between the weather and Citi Bike ridership
- 33. Modeling the relationship between the weather and Citi Bike ridership • Daily ridership data from Citi Bike • Daily weather data from National Climatic Data Center: temperature, precipitation, snow depth • Devise and calibrate model in R
- 34. Modeling the relationship between the weather and Citi Bike ridership
- 36. Calibration in R • Uses nlsLM() function from minpack.lm package for Levenberg– Marquardt algorithm to minimize nonlinear squared error https://gist.github.com/toddwschneider/bac3350f84b2ff99969d
- 37. Model Results
- 38. Airport traffic • Question: how long will my taxi take to get to the airport? • LGA, JFK, and EWR are each their own census tracts • Get all trips that dropped off in one of those tracts • Calculate travel times from neighborhoods to airports
- 39. Airport traffic
- 40. More fun stuff in the full posts • On the realism of Die Hard 3 • Relationship between age, gender, and cycling speed • Neighborhoods with most nightlife • East Hampton privacy concerns • What time do investment bankers arrive at work?
- 41. “Medium data” analysis tips
- 42. What is “medium data”? No clear answer, but my rough thinking: • Tiny: fits in spreadsheet • Small: doesn’t fit in spreadsheet, but fits in RAM • Medium: too big for RAM, but fits on local hard disk • Big: too big for local disk, has to be distributed across many nodes
- 43. Use the right tool for the job My personal toolkit (yours may vary!): • PostgreSQL for storing and aggregating data. Geospatial calculations with PostGIS extension • R for modeling and plotting • Command line tools for looping through files, loading data, text processing on input data with sed, awk, etc. • Ruby for making API calls, scraping websites, running web servers, and sometimes using local rails apps to organize relational data • JavaScript for interactivity on the web
- 44. R + PostgresSQL • The R ↔ Postgres link is invaluable! Use R and Postgres for the things they’re respectively best at • Postgres: persisting data in tables, rote number crunching • R: calibrating models, plotting • RPostgreSQL package allows querying Postgres from within R
- 45. Tip: pre-aggregate • Think about how you’re going to access the data, and consider creating intermediate aggregated tables which can be used as building blocks for later analysis • Example: number of taxi trips grouped by pickup census tract and date/time truncated to the hour • Resulting table is only 30 million rows, easier to work with than full trips table, and can still answer lots of interesting questions
- 46. Pre-aggregating example CREATE TABLE hourly_pickups AS SELECT date_trunc('hour', pickup_datetime) AS pickup_hour, cab_type_id, pickup_nyct2010_gid, COUNT(*) FROM trips WHERE pickup_nyct2010_gid IS NOT NULL GROUP BY pickup_hour, cab_type_id, pickup_nyct2010_gid; https://github.com/toddwschneider/nyc-taxi-data/blob/master/analysis/prepare_analysis.sql#L30-L38
- 47. How to get people to read your work • It has to be interesting. If you’re not excited, probably nobody else is either • Most people are distracted, and they read things in “fast scroll” mode. Optimize for them • The questions you ask are more important than the methods you use to answer them
- 48. Specific tips • Write in short paragraphs with straightforward language • Use plenty of section headers • Good ratio of pictures to text • Avoid the dreaded “wall of text”
- 49. Above all… • Have fun! • Keep an inquisitive mind. Observe stuff happening around you, ask questions about it, try to answer those questions