Skip to content

Latest commit

 

History

History
127 lines (101 loc) · 8.7 KB

PRESENTATION_TRANSCRIPT.md

File metadata and controls

127 lines (101 loc) · 8.7 KB
Raw

Presentation

Project - Citi Bike Information Visualization

Content

Slide 1 - Title

Slide 2 - Why chose this project

  • Interesting: Sharing economy is popular in recent years. As one of typical sharing economy, Bicycles are popular vehicles in our daily life, shared bicycles perfectly solves the problem of the last one mile in the city for people. It is also a healthy and eco-friendly means of transportation.All of group members like to use sharing-bike.
  • Useful: For project itself, sharing-bike trip data has enough dimensions. We want to analyze different data to find potential factors. And help bike-sharing company make decisions.
  • Important: Company will know when and where to set up more share bikes. And bike users know where they can find a share bike. It helps company to offer a better services and maximize the use efficiency of every bike.

Slide 3 - Target Users

  • Citi bike-sharing company. If they want to expand their business in other cities or want to set more new stations, our website can help them to know more factors.
  • People who want to know information of bike-sharing. User can get some information from our website that help them use bike.
  • Competitor of Citi bike-sharing. From our website, they can get some useful information and laws to help them make business decisions.

Slide 4 - Motivation and Goal

  • I have been to NYC for vacation trip. And I found there lots of bike-sharing stations around my hotel. But all of them have no available bikes for most of time. The contradiction is there are hundreds of bike-sharing stations at NYC but I still cannot borrow bikes at anytime and anywhere. So we decided to find potential factors that may influence orders and station's location by doing data visualization.

Slide 5 - What others have done in this topic and what we do differently

  • Relevant website
    • Other websites with the same topics offers a lot of statistics charts and spatial analysis. We referred to some of charts such as returning bikes and borrowing bikes (per hour) statistics bar chart.
    • They focuses on statistics and analysis. Their target users are tend to data analysts.
  • What we do
    • We used simple statistics charts to tell a story about sharing bikes, showing the temporal and spatial variation of stations and the factors
    • Our charts are responsive and interactive. We optimized the visual queries and user interaction by multiple methods such as pop out effect, coupling effects and appropriate color scheme.

Slide 6 - Data

  • Citi Bike Trip Histories
    • The dataset is about trip information and is summarized for each month from 2013 to now
    • Includes Start Time and Date, End Time and Date, Start Station Name, Start Station Name, Station Lat/Long, Station ID, User Year of Birth, etc
    • We downloaded trip data of June from 2013 to 2018
    • We processed the original data and generated Station GeoJSON data of 6 years, total number of borrowed bikes per hour of each station for June of each year, total number of returned bikes per hour of each station for June of each year and user age summary.
    • Source: https://www.citibikenyc.com/system-data
  • NYC Facilities
    • The New York City facilities data is offered by NYC Capital Planning Platform
    • includes various types of facilities such as core infrastructure, historical sites and parks
    • The format of the data is GeoJSON
    • We used historical sites, education and infrastructure datasets to show the surroundings of top 13 popular stations.
    • Source: https://capitalplanning.nyc.gov/map/facilities#10/40.7128/-74.0807
  • Precipitation Data

Slide 7 - Station Map (Statistics Analysis)

  • Two statistics bar charts
    • Total number of borrowed bikes per hour of June in a year for one station
    • Total number of returned bikes per hour of June in a year for one station
  • Select any station of one year on map to see the statistics
    • Which station is popular and which is not
    • When a user might get a bike, while when might not
  • Responsive
    • The chart will adjust based on the size of the window
  • Interactive
    • Users select stations on the map and see the transition of charts
    • Hover a bar and see the number of bikes
  • Coupling effect
    • Two bar charts shows pop out effect together when a user hover a bar

Slide 8 - Main chart design/layout

  • Mapbox, ant-design library
  • Bar chart / Multi-line chart

Slide 9 - Stations Map (Distribution Variation)

  • One statistics line chart
    • Average activities of total stations per hour
  • Table of top 5 neighborhoods with most new installed stations

Slide 10 - Stations Map (Distribution trend)

  • From 2013 to 2018, there are 439 new installed stations
  • Most of them are distributed in north and east, which are in the neighborhoods like upper town and Brooklyn.
  • However, we found some of these new installed stations are inactive in our statistics. We would like to explore why popular stations are popular and to see if these new installed stations are necessary.

Slide 11 - Station infrastructure analysis

  • Base on our story, we located nearby station on map. In the same time, we found this station is one of top10 popular stations around NYC. By doing these, we found there are many malls and hotels around this station. So we thought the surroundings of station is also a key factor which may influence number of orders. Thus we got top 20 popular stations data and their nearby infrastructures data.

Slide 12 - Station infrastructure analysis 2

  • Our chart can directly compare popular stations. As we can see, those popular stations always surrounded by malls, parks, hotels, transportations, big apartment and so on. For example, station located at south entrance of central park is one of top 20 popular stations. There lots of users came and left central park by riding sharing-bikes.
  • D3.map,Bootstrap,function.
  • Zoom and brushed can help user read long period data more easily.

Slide 13 - Station weather analysis

  • Based on our experiences, weather is a important factor that may influence orders. Because, we all don't want to ride a bike at rainy data or snow day. So we got some data from New York open data site. Weather data includes maximum and minimum temperatures, perception and other relevant data. Apart from weather, seasons is also a key factor that may influence orders.
  • According to our chart, we found there more orders at spring, summer and autumn. And there are less orders on rainy day.

Slide 14 - Station user analysis

  • We think different stations have different features. For example, stations near school and universities have more young users and stations near park have more old users. We try to find some user portraits base on their age data. Thus we got top 10 stations age data in 6 years.
  • Base on our data, we didn't find laws.

Slide 15 - Who did what

  • Ziwei Yuan
    • Built the framework of mapbox including loading mapbox, adding legend and adding layers
    • Implemented components Interaction between mapbox and the side area by adding a service
    • Added pop out effect when hovering or clicking a station on the map
    • Drew two responsive and interactive d3 bar charts (return bikes and borrow bikes) in the map view
  • Tong Lyu
    • Designed the layout and build entire framework of the website.
    • Created the line chart of "Distribution Variation".
    • Create a table and bar chart count of top 5 neighborhoods with most new installed stations, and highlight the neighborhood on the map.
    • Participated in integrating mapbox to display stations and listen to mouse hover events.
  • Tianyang Li
    • Got raw data and format these data into csv, json and relevant useful files.
    • Created d3.map with top20 popular/not popular stations data combine zoom function to analyze infrastructure factors.
    • Merged pie charts, line chart and map chart into same page. Also designed website.
  • Jing Zhang
    • Analyzed top 20 popular/not popular stations data and find relevant factors.
    • Created line chart with zoom and brush function.
    • Created donut charts analyze relationship between orders and user's age.