I am graphing the meteo temperature of bordeaux the last 90 days.
Each column is a day and each row is an hour. Each circle is the temperature.
The size of the circle is the temperature value between the maximum and minimum of the period, this last 90 days.
To my mind it is beautilfull so i don’t want to add any extra text. I think i will add it on the mouse over.
Minimal data visualisation have many qualities:
- we really focus on our form /pattern recognition brain.
- we are sure the reader is a specialist: he has enough extra info in his brain to add value to the graph.
- it is beautiful so the reader want to watch it and spend time watching it.
The next step is to add variation to the data pre-process. For example scale the circle size only on a day basis so we the warmer moment of the day and so outliers: maybe some days warmer moment is not at the same hours.
I will maybe add hover on the circles or days for extra text information.
I wonder to create a small time series dataviz plateform on this basis, with elasticsearch of influxdb in backend. It will be a polyglot database backend.