API Endpoints (/api/v1/...
)
-
/poverty/map
(?year=2014
) (state geo.json data with any coloring/shading options applied (maybe a color intensity?) -
/poverty/markers
(?year=2014
) (any individual points or markers we want on/in the states, possible year filter) -
/poverty
(general data set information (data set name, updated_at, avail. years, etc)) -
/poverty/state/tweets
(last 50 tweets within predefined poverty twitter hashtags) OR/poverty/tweets?state=XX
-
/poverty/state/tweets/sentiment
(sentiment scale score of last 2000 tweets) -
/poverty/state/data
(?year-2014
) (more chosen state data, for example actual numbers/%'s, that could be overlaid on state) -
/poverty/state/images
(if we wanted to pull in instagram photos) -
/overview
(contains an overview, list of available datasets/endpoints (poverty, drug use, etc), etc.)
Mapping:
- Sample D3 based map
- http://bl.ocks.org/mbostock/4699541
- Seems like with some customization this could be a good basis for our main map
- Its also cool that its based on D3 which it sounded like we all kind of wanted to play with a little?
State Data Nesting:
- Seems like this could go one of two ways, either state specific routes /CO/, /NY/, etc, or a state code could just be passed through as a filter to a more generic route /tweets?state=CO&year=2014 (which may be better the more I think about it... seems like this might allow you to minimize the complexity and number of api requests allowing more filtering to be done on the front-end)
- In the same vein, I'd guess sentiment is probably just saved as a tweets data attribute
-->>
{ data: { sentimentality: 0.75, sentimentailty_verbose: 'Slightly Negative', pulled_at: 'dateTimeObj', tweets: {[ {id: 1, user: sdfsdf, user_prof_link: 'www.twitter.com/sdfsdf' user_img: 'http://ssdfsdf.com', text: 'sdfsdfsdf', time: 'dateTime', trigger_hashtags: ['#poverty', '#poor']}, {...}, {...} ]} } }
Main Data/Reports
- Need to decide on one sample report to focus on to start with
- If we made available years clear (i'm thinking something in the /poverty endpoint), i think we could process data on an annual basis without too much extra work, I'm guessing it would just be providing a filter option to the endpoints above where I included a filter parameter
- the /poverty/state/data endpoint was for storing things like actual reporting statistics so we could overlay those on the map... or something like that...