FairElection - EF2019

Small website showing the various results of the Swiss 2019.

Editing

This website is built with Nuxt.js, and thus Vue.js. Please refer to their relative documentations for general information:

• Vue.js docs
• Nuxt.js docs

For the sake of convenience, all typical folders usually found at root level in a Nuxt.js project have been moved to src in this project. To start using the project, install the required dependencies. Browse to the root folder of the project and use the regular installation command:

# if using npm
$ npm install
# if using yarn
$ yarn install

As this project is meant to be deployed statically on a simple webserver, all changes require to rebuild the whole project. To do so, browse to the root folder of the project and use the build generate command. This will create a dist folder at the root folder of your project: this folder can then be served directly through a webserver.

# if using npm
$ npm run generate
# if using yarn
$ yarn generate

Development

To save some precious time during development, use the command dev to enable hot-reload along with a local webserver. Any modification of the source code will thus be repercuted directly in your browser as you go.

# if using npm
$ npm run dev
# if using yarn
$ yarn dev

Translations

This website relies on Nuxt i18n (and thus, Vue i18n) to handle internationalization.

• Vue i18n docs
• Nuxt i18n docs

For convenience, the language files are located in src/assets/lang. Please make sure to regenerate the website if you change them.

Data

All data used by the website can be found under src/assets/data.

• cantons.json: contains the list of Swiss cantons, and their name in French and German;
• elected_candidates.json: contains all candidates and their relative information (demographics, election by scenario, etc);
• parties.json: contains all the parties represented by the candidates;
• repartitions.js: contains the three dimensions used by the scenarios (age, gender, location) along with their possible values;
• scenarios.js: contains all the scenarios;

While most files are self-explanatory and fairly easy to manipulate, elected_candidates.json is the most complicated file to generate, as it relies on external data. Under src/assets/data-processor, you will find the following items:

• Dist_*: various folders containing the results of the balanced election model for all scenarios, one for each Swiss canton;
• Votes: a folder containing the final results of the 2019 elections;
• candidates.csv: an external data file containing various demographic data regarding candidates;
• convert-base.js: a node script to generate the scenario_base_towork.json based on the two files mentioned above;
• convert.js: a node script to generate elected_candidates.json based the cantons' data and scenario_base_towork.json;
• package.json: a NPM file to install required dependencies to process data;
• scenario_base_towork.json: a data file containing all the elected candidates of the 2019 elections.

Please note that elected_candidates.json contains only those candidates who are elected in at least one scenario of the simulation, as to avoid over 3K of unused candidates' data. If you intend to rebuild the file, you will have to install a local library called csvtojson to parse CSV files in Node.js. To do so, browse to src/assets/data-processor and use the following command:

# if using npm
$ npm install
# if using yarn
$ yarn install

Then, you should be able to generate the results files. In the unlikely case where you need to regenerate the scenario_base_towork.json file, use the following command from the same folder:

$ node convert-base.js

You will then need to manually fix candidates that could not be matched against candidates.csv. To do so, search for "null" in the generated document: these candidates are either not found at all (suggestions property is empty) or found several times (suggestions property has more than one value). Find the relevant entry in the candidates.csv file and fill the "null" fields.

In the case you need to regenerate the elected_candidates.json file, use the following command:

$ node convert.js

This should generate various files, including a new elected_candidates.json in the folder. You can then replace the one in src/assets/data and regenerate the website.