Temperature in Different Cities
This visualization presents a small multiples area chart tracking temperature readings across seven cities over one week in 2015, using data from the Sense Your City project. The design intentionally demonstrates an ineffective approach to data visualization: color encoding is arbitrary, the area under each curve holds no quantitative meaning, and comparing values across the separate faceted charts proves difficult. Built with Vega-Lite 2.0 and D3 v4, the code constructs the specification programmatically, loading CSV data via `d3.csv` and rendering the output through `vegaEmbed`. Each city occupies its own row, with temporal timestamps on the x-axis and temperature in degrees Celsius on the y-axis, while grid lines are disabled to simplify the display.
AI-generated descriptionSmall multiples area chart for one week in 2015.
This is an example of how not to visualize data:
- Color is meaningless
- Area under the curve is meaningless
- It's difficult to compare across area charts
Draws from
- News Source by Age Grouped Bar Chart for Vega-Lite setup.
- Vega-Lite: Trellis Area Example
Also related - Vega-Lite GitHub Issue: Specify Label Width for Facet
Data: a week of rich sensor data collected by the "Sense Your City" project.
Contains:
- Hourly readings for one week
- Different cities: San Francisco, Bangalore, Boston, Geneva, Rio de Janeiro, Shanghai, Singapore
- Various mertics from sensors: temperature, light, air quality, sound, humidity, dust
The data was collected from the (now defunct) Data Canvas - Sense Your City Project. The data was filtered using the data processing script found in GitHub: curran/data/senseYourCity.