Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Collaboration on Cab/Rideshare Data Visualization #267

Open
AndrewTBurks opened this issue Nov 10, 2016 · 6 comments
Open

Collaboration on Cab/Rideshare Data Visualization #267

AndrewTBurks opened this issue Nov 10, 2016 · 6 comments

Comments

@AndrewTBurks
Copy link

Hi all,

My name is Andrew Burks from UIC, and for my data visualization class this semester, I am working with Tom Schenk to visualize Cab/Rideshare data using OpenGrid. He advised me to create an issue on GitHub to get in contact with the entire team, in order to make communication easier and to initially guide me in regards to working with OpenGrid.

As of now, I have a few hand-drawn visualization prototypes for the Cab/Rideshare data which I would like feedback on before I move forward:

Cab Visualization Prototypes

Please let me know if I need to clarify anything.

Thank you in advance for your help.

-Andrew Burks

@tomschenkjr
Copy link
Contributor

@AndrewTBurks - thanks! We are working on releasing taxi data, but I'd encourage you to look at the DIVVY bike trip data (JSON), which has a similar structure (an origin lat/long and destination lat/long). We will look at your sketches as well. (cc'ing @abemmanuel who is at CDOT).

Just to clarify for @Chicago/open-grid-team - the objective is to allow OpenGrid to query and visualize transit data, such as taxi trips, bike trips, bus, etc. Taxi trips (and DIVVY trips) serve as a good example. This is different from current capabilities because trip data often needs the show the relationship between at least 2 points.

Also cc'ing @WillEngler since this might spark thinking around Plenario and transportation data.

@AndrewTBurks
Copy link
Author

AndrewTBurks commented Nov 23, 2016

Hi all,

I have been working with the cab data and I have some progress using a separate Leaflet map. This work so far goes off the idea of querying origin data within a selection area and viewing destination data outside of the area. (Note: The color schemes will be adjusted. Also, the area selection isn't very intuitive, the Leaflet plugin I am using isn't the most optimal choice for this task.)

Link: Cab Data Vis

Let me know any feedback you have, as I haven't heard back much on this project.

Thank you,
Andrew

@stevevance
Copy link

stevevance commented Nov 26, 2016

@AndrewTBurks Upon page load a box appears which has grids that are symbolized. If I move the map under the grid, the symbolization disappears. If I change the grid dimensions, the symbolization disappears. I don't know what the symbolization shows, and I don't know how to manipulate the map.

@AndrewTBurks
Copy link
Author

Hi @stevevance,

Thank you very much for your feedback!

Sorry if it was a bit unclear, it was still in early stages of development. The idea was that the user could query trips with origins within the selection box, and the map would show the origins within the selection (encoded using the grid heatmap), and the destinations throughout Chicago would be shown by census tract. The amount of cab trips is encoded by the color of the heatmap (blue grid) and choropleth (red census tracts). The symbols disappearing was a new request to the database being made (automatically when the selection area was moved) and the selection grid and cesnsus tracts would be updated when the query completed.

I will be working on this a lot and I will have another update within the next day or two.

@AndrewTBurks
Copy link
Author

Hello everyone,

As it is nearing the end of the semester, I am wrapping up the current version of this project. I was hoping to get some feedback on the work I have done so far. This is the current state of the tool, with an explanation of the different operations:

CabViz

There are two modes to the visualization.

Query Mode

The first mode is the "Query" mode, which will allow the user to query pick-ups or drop-offs within the selected location. For example, if you want to look at traffic from around Millenium Park, the selection box can be placed over the Millenium Park and the button "Query Pickups" or "Query Dropoffs" can be clicked to retrieve trips from this area to the rest of the city, or from the rest of the city to this area, respectively.

The traffic within the selection is encoded as a grid heatmap. The traffic in the rest of the city is colored by census tract.

(Note: The number of entries that the API call retrieves is small currently, so the data may look a bit strange. Also, due to the position obfuscation, at some scales the data looks very spotty)

Time Analysis Mode

The second mode is the "Time Analysis" mode. The idea behind this mode is showing how traffic changes within a time window, analyzing the traffic before and after an event of interest. In the tool you see now, the time analysis is done over the 2 day period:

start = "2015-01-01T12:00:00", end = "2015-01-03T12:00:00"

with the event at

event = "2015-01-02T12:00:00"

The color of the census tracts corresponds to the % change in the volume of trips before vs. after the event, while the opacity encodes the total volume of trips (more opaque census tracts have had more traffic, regardless of the %change). With these encodings, an opaque census tract which has a strong color will mean that a high volume of trips had a dramatic change from before to after an event. When a census tract is clicked on, a graph is drawn showing the trips, separated by pickup and dropoff. The graph which is drawn also shows the time of the event as a yellow highlighted bar. The data in the graph is taken at 15 minute time intervals (consistent with how the trip start/end times were rounded in the data) and the line which is shown is interpolated. I used a monotoneX interpolation to preserve the data points, where a Basis interpolation could pull the line off of the actual data points while smoothing.

@stevevance
Copy link

I tried your example, Millennium Park. It takes a tiny bit of getting used to how to move the map to fit under the red box, but I got it.

This map is showing the search area of pickups (a couple hundred feet in all directions of Millennium Park) and showing the destinations of those passengers. The darker the brown, the more trips that ended in that Census tract.

A lot of people are getting picked up on the blocks surrounding Millennium Park and going to the Loop.

screenshot 2016-12-06 15 32 18

In those blocks surrounding Millennium Park, most people are starting their taxi ride in a (purple) box that includes the intersection of Monroe Street and Columbus Drive; the Art Institute of Chicago is at one corner, and Maggie Daley Park is at another corner (well, it's been there since December 2014).

screenshot 2016-12-06 15 34 14

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

4 participants