Visualizations and data procesing for the Open Data Census.
This also includes various ancillary information providing an overview of what is happening with release of open government data around the world (and initiatives related to it).
Read the Project README and Overview.
Census is being performed at 2 different levels:
- Country level
- City / regional level
The set of datasets asked about for each level are different but the questions asked per dataset are very similar.
We use google spreadsheets as a backend for the Open Data Census. Data is stored in 3 different spreadsheets:
- Dataset and Questions spreadsheet - this list the datasets for the country and city census as well as the questions asked for each dataset
- Country spreadsheet
- City and Region spreadsheet
For historical reasons the raw data from the Country and City census arrive in different forms. We aim to get
We attempt to have a common form to the results spreadsheet.
- Submissions - this is raw data from form submissions
- Normalized - raw submissions data but in normalized form (standard headings)
- Reviewed (not yet used) - reviewed and consolidated data
Process
- Submit via google form
- Arrive in submissions tab
- Copy to the Normalized tab (which has standard columns)
- Review (soon)
Process
-
Submit via our custom JS form. Data is still submitted to a google spreadsheet with a google form. However we do not use Google's generated form but just submit ourselves using JS. The form therefore only has 1 field "Data") and we store results of form as a JSON blob in this field.
-
Normalized sheet - this is produced from submissions sheet by running the unpack script (go to Open Data Census menu => Unpack)
- uses this app script https://script.google.com/a/macros/okfn.org/d/M2ShE7x1mR_sRh_6fT27vYhyeNKfa_QZS/edit
- Reviewed
The app is a simple Express NodeJS app deployed on Heroku.
Install
To install do the following:
1. Get the code and init the submodules (for recline vendor library)
git clone https://github.com/okfn/opendatacensus
git submodule init
git submodule update
2. Install node dependencies
cd opendatacensus
npm install .
3. Run the app
node app.js
4. Should now be running at <http://localhost:5000>
Running Tests
Install mocha (see devDependencies in package.json) then do:
mocha test/
note we run off live data yet some expected values are hard-coded. Thus tests are likely to break as underlying data changes. We should fix this soon by running off mock data.
Deployment
App is deployed on Heroku as `opendatacensus`. Do the usual.