LA Grub Grabber (LAGG) is Live

Project Website

Eminent Domain Final Presentation

PROJECT DESCRIPTION

Finding a restaurant in Los Angeles (LA) isn’t a challenge, but finding a good one that is accessible by bus or car can be. The LA Grub Grabber (LAGG) tries to the fill the void by providing you with the tools to find a great place to chow down. 

Click on the map and a 500 meter radius of that point shows all the restaurants contained within it by accessing the Citysearch API. Additionally all of the bus stops within the 500 mile radius are located and their associated bus lines are displayed. Each bus stop (on the lines that fall within the 500 meter radius of the original point) within a 30-minute time period (factoring in walking time to the station and the bus schedule) is displayed. Each one of those bus stops has another 500 meter radius displaying all of the restaurants within it. This functionality allows users to see what restaurants are along their chosen bus route.

The number of restaurants and restaurant density by zip code allows users to choose bus routes or restaurant locations with an abundance of restaurants, should they choose to do so.

If users, or their friends, prefer to drive, we show a layer of all City of LA parking lots.

Users can also insert their favorite restaurants by using the Google Form and geocoding it.

USER INTERACTION

(1) Google Form

(2) Clickable Layers

(3) Click on the map and get restaurants and bus routes

CUSTOM FUNCTIONS

Bus Network Algorithm:

•Transformed  Metro GTFS data into the table structures we needed and
created Fusion Tables services to provide trip planning functionality.
•We use lat/lon to look up stops within max walk distance
•Get departure times and trip IDs for nearby stops within time threshold
•Find arrival times and stop IDs within the time threshold for the
associated trip IDs
•This gives us a set of stops that are reachable within the max time threshold, with a variable walk distance around each up to the max time, or max walk distance (whichever is lower).  We add a marker for each stop.

Restaurant Finding Algorithm:

•Easy way: Do a radius search for each stop.  But this is very slow.
•So we compute a bounding box for the stops and do a single category/box search on CityGrid, giving us a large result set.
•We then iterate through each CityGrid result and see whether it is within the max walk radius of any of the stops.  If so we find the shortest path (walk + bus) to the location and then check whether the total trip time is within the threshold.
•If the location is within the max trip time threshold and max walk distance, we add it to the map.

CUSTOM LAYERS

Restaurant Density by Zip Code / Number of Restaurants

City of LA Parking Lots

EVALUATION

Ideally we would have been able to use the Metro API, but this caused us a lot problems. We would also have liked to perform some basic statistics using JStat (or another program), since this would have provided an interesting analytical component.

DIAGRAMS


DIVISION OF LABOR

David Benoff wrote the tricky code and had great UI ideas…David Peterson grabbed the data, created layers, and learned some programming tricks from DB.

Final Project/Presentation

Refocused Project Theme:

After experimenting with importing Google Fusions Tables and the Yelp API for the mid-term project, Eminent Domain (ED)  is refining its take on access in Los Angeles, by looking at the relationship between business density and travel options. Business density will be determined by a business density layer (possibly by census tract), and the Yelp API click event which displays all the business in a 1 mile radius of wherever the user clicks on the map.

The  transportation piece will address both drivers and transit users. City of LA parking lots will show where there is available parking near high business density (or low business density for that matter) areas should the person choose to drive. Should the person choose transit, the Metro API will load all the bus routes and allow the person to select the most appropriate one.

Google Forms will allow the user to plot points on the map for various purposes. For example, if the user is interested in identifying all of the businesses in a certain area, the user can insert the address, and have a marker identify the location. Then, using the Yelp API, she can click near the marker she just generated and display all of the businesses in a given area. Similarly, when choosing bus routes, the user could input the starting and ending points of the trip and select the most appropriate routes.

The wire frame of the website is here:

Eminent_Domain_Final_wireframe

Current status: we are currently putting this together and continuing to gather data.

Problems: No issues for the moment

Eminent Domain Mid-Term Project

Project website: http://dbenoffa.bol.ucla.edu/midterm.html

Project Team: Eminent Domain (David Benoff + David Peterson)

Project topic: Access in Los Angeles

Description:

We attempt an interactive approach at measuring access in Los Angeles County with a Google Maps mash-up incorporating Yelp, City of Los Angeles Parking lot data, and year 2000 population density by Census Tract. We define access as the relative measure of useful (to the user) opportunities within a given area. For example, a destination that had only 1 activity (e.g. a McDonald’s) would have low access and a destination that had many activities (e.g. a McDonald’s, a chiropractor, supermarket, and a coffee shop) would have higher access.

The value attributed to these activities is determined by the user, which is why we hope users will find this website helpful when thinking about housing location decisions, where to shop, and whether to drive or take the bus to accomplish their activity.

Functionalities:

  • Parking: layers can be toggled on/off; info windows

  • Population Density by Census Tract: this was rendered in Google Fusion Tables API. An info window pops up with a click event on the census tract, displaying information about the tract.

  • Yelp Click Event: clicking anywhere on the map will render markers for businesses within a 1-mile radius. Clicking on each marker will give you the name of the business entity and its category (e.g. Z Gallerie Outlet, home décor)

Who is it for, why is it useful, how are you implementing it:

The website is intended to be used by individuals, students, policymakers, and anyone trying to make a decision about where they should go to accomplish an activity and what transportation mode they should use (solo drive, bus, rail, walk, bike, etc.).

Yelp: Knowing what kinds of business are clustered in any given destination can help you decide which cluster you would prefer to go to. If the user is looking to accomplish three tasks, the Yelp tool can help accomplish that by informing the user of what business are located nearby.

City of LA Parking Lots: Curb parking is typically scarce in areas with high access and knowing if a city parking lot is available can influence what mode of travel will be used. Real time parking information would be more useful, but this is not currently available. In the future we intend to add bus and subway routes, as well as a bicycle layer (if available).

Population Density: This layer was added to test whether or not there was a correlation between population density, a high degree of transportation infrastructure coverage (parking lots, transit routes, bike routes, etc.), and activities. Developing this further could lead to interesting policy implications, real estate and infrastructure investments, and other unforeseen possibilities.

Description of roles and what each team member did:

David B: Final website design; Yelp API, documentation

David P: Data; Google Fusion Tables layers; documentation

Challenges:

  • Fusion Tables Layer: getting the info windows to display properly when 2 Fusion table/KML layers were being displayed. The problem was that the info windows for only one data would display.
  • Calculating population density: we obtained land area and population for census tract, but there seemed to be some problems with the data (population seemed too high or too low in many cases), and we contacted the UCLA data librarian to assist us. The density map looks correct, after comparing it to other density maps of LA County. The problem can be seen in the range for the population density quintiles.

Additional features to implement for the Final:

Some of the following will be added:

  • LA Metro API
  • LA County Bicycle Map Layer
  • Traffic Data
  • Statistical Analysis tools (JStat)

Revised Scope and “New” Team

The Eminent Domain team is now down to two members: David Benoff and David Peterson. The previous two members (Lindsey Hilde and Melissa Kelley) are no longer registered for the class. Given this shake-up we had to re-group and submit this post this morning.

The statistical component of the website will be eliminated given the apparent complexity of executing the task. Jstat has potential, but integrating it into our website will be difficult. Instead we have decided to focus on the following:

  • Incorporating influenza data from the CDC through Google Fusion Tables, and making a call to the Fusion Tables API from the browser to populate a map. This task could require rendering a Google Earth layer outlining the CDC regional boundaries/polygons.
  • Incorporate a real-time data element from the CDC, much like we have incorporated Flickr and Yelp in class assignments.

Data sources are here.

We will post revised Midterm and Final project goals after class on Monday (April 25).

Project Milestones

Midterm: Beta Launch

On Monday, May 2, 2011 of Week 6, our team will have completed the following milestones:

  • Completion of a functioning beta version of our website built using Google Map v3 API hosted on Google Fusion Tables.
  • Incorporation of at least two data feeds (see Datasets above)
  • A blog entry about the Eminent Domain team, our goals, and experience
  • A blog entry launching the beta version of our website.
  • A short, 10-15 minute, presentation about our beta site involving all team members. Feedback we gather from our classmates and instructors will be instrumental in helping us improve our site for final launch at the end of the term.

The blog entry launching our site will serve as our main platform for communicating the functionality of our site to potential users. This entry will explain the process we undertook as a team in order to build a functioning site in response to CDC Flu App Challenge. The roles of each Eminent Domain team member will be identified.  We will describe our topic, as put forth by the CDC Flu App Challenge, and walk the user through a description of our site’s features and functionalities. We will create and display diagrams, flowcharts, and sketches depicting the flow guiding our site’s functions as well as the planning process we undertook in the five weeks leading up to the launch. We will also discuss our successes and challenges over these five weeks. Finally, we will describe our goals for the final website to be delivered on May 27th (due date for CDC Flu App Challenge entries) and discuss in more detail the steps we will take as a team to realize thee goals in a timely and efficient manner.

Final: Full Launch

On May 27, 2011, our team will have completed the following milestones:

  • Addition of at least one more data feed that allows the user to draw out more detailed statistical analysis than previously offered in the beta version.
  • Improve the user interface of our site, based upon feedback from peers and instructors
  • Use of a consistent color palette that is artistically complementary and user friendly.
  • Improve and solve any back end programming issues involving Fusion Tables and layers. If necessary define boundaries for data lacking coordinates

On June 6, 2011 our team will have completed our final milestones:

  • A blog entry announcing the final launch of our site
  • A short, 10-15 minute, presentation about our beta site involving all team members.

Proposal Topic & Site Functionality

Public Health and Urban Planning have a long history, and we hope to demonstrate the synergies between these two fields through innovative uses of information technology. The Centers of Disease Control and Prevention (CDC) provides some highlights on this intertwined history (source: http://www.cdc.gov/mmwr/preview/mmwrhtml/su5502a12.htm):

During the 19th and early 20th centuries, the synergies between urban planning and public health were evident in at least three areas: creation of green space to promote physical activity, social integration, and better mental health; prevention of infectious diseases through community infrastructure, such as drinking water and sewage systems; and protection of persons from hazardous industrial exposures and injury risks through land-use and zoning ordinances. During the middle of the 20th century, the disciplines drifted apart, to a certain extent because of their success in limiting health and safety risks caused by inappropriate mixing of land uses.

The disciplines recently have begun to reintegrate. During the last 20 years, shared concerns have included transportation planning to improve air quality, encourage physical activity, prevent injuries, and promote wellness. In addition, some original crossover ideas, such as the potential for parks and recreational facilities to contribute to physical activity and mental health, have reemerged. Relatively recently, urban planning has focused on the effects of community design on energy use and greenhouse gas emissions to affect the growing public health concern of climate change. Finally, emergency preparedness (e.g., community infrastructure assurance, evacuation planning) and access to health care (e.g., assurance of accessibility and adequacy of facilities) are topics important to both disciplines.

Our project intends to employ spatial statistical techniques to analyze influenza data provided by the CDC. We hope this information will be useful to urban planners, geographers, and public health professionals.

(2) Functionality

Users will be able to select CDC data (and possibly other data, time permitting), which will be hosted on Google Fusion Tables. After selecting the data, the user will choose one of a number of statistical techniques to be performed on the data. We will first attempt to provide measures of central tendency, and if successful, we will provide more advanced statistics. The statistical operations will be executed in the browser, and will render the resulting images on a map. In addition to the statistics, the CDC features offers several RSS feeds (e.g. podcasts, influenza updates), which will feature on the website and provide real time information.

Project Datasets and Sources

Project Datasets/Data Sources

The mission of the CDC Flu App Challenge is to promote healthy behavior for flu prevention by making it easier to communicate critical information about the flu and its impact.  In order to raise awareness about influenza and/or to educate consumers on ways to prevent and treat it, our website will utilize several data sources.  Specially, data will be obtained directly from the CDC, general health sources and general population sources.  Technological applications will include several mapping and visualization tools to help promote best practices for influenza health and risk communication.

Mapping/Visualization APIs

Google Maps API: http://code.google.com/apis/maps/documentation/javascript/basics.html

Google Fusion Tables API: http://code.google.com/apis/fusiontables/docs/sample_code.html#ftl

Google Visualization API: http://code.google.com/apis/visualization/documentation/gallery.htm

Google Charts API: http://code.google.com/apis/chart/docs/making_charts.html
CDC Flu Data

CDC contest datasets:

Influenza Vaccination Estimates
http://www.cdc.gov/flu/professionals/vaccination/reporti1011/resources/2010-11_Coverage.xls
XML source of the Weekly Flu Activity Report
http://www.cdc.gov/flu/weekly/flureport.xml
RSS Feed of Influenza pages through content syndication
http://t.cdc.gov/feed.aspx?tpc=26829&days=90
RSS Feed of Influenza updates
http://www2c.cdc.gov/podcasts/createrss.asp?t=r&c=20
RSS Feed of Influenza podcasts
http://www2c.cdc.gov/podcasts/searchandcreaterss.asp?topic=flu
RSS Feed of CDC Features pages through content syndication
http://t.cdc.gov/feed.aspx?tpc=26816&fromdate=1/1/2011
JSON Feed of Influenza pages through content syndication
http://t.cdc.gov/feed.aspx?tpc=26829&days=90&fmt=json
JSON Feed of CDC Features pages through content syndication
http://t.cdc.gov/feed.aspx?tpc=26816&fromdate=1/1/2011&fmt=json

CDC Seasonal Flu Activity: http://www.cdc.gov/flu/weekly/fluactivitysurv.htm
Health Data

Health Indicator Warehouse API: http://www.healthindicators.gov/Developers/Overview

Medline Plus API: http://www.nlm.nih.gov/medlineplus/webservices.html

Health Data Interactive: http://www.cdc.gov/nchs/hdi.htm

Distribute: http://isdsdistribute.org/

WHO FluNet: http://www.who.int/csr/disease/influenza/influenzanetwork/flunet/en/
Other Population Data (demographics, geographic boundaries, etc.)
U.S. Census TIGER/Line: http://www.census.gov/geo/www/tiger/
School information: http://www.greatschools.org/api/registration.page

Good and Bad Health Data Example Sites

http://www.google.org/flutrends/

http://healthmap.org/swineflu/ (here is link to with information about the site http://healthmap.org/about/)

http://www.healthindicators.gov/Indicators/Flu-vaccination-adults_119/California_6/Profile/Data

http://www.cdph.ca.gov/data/statistics/Pages/H1N1Data.aspx

http://www.cdc.gov/flu/weekly/