This is an online application that presents backdated meteorological data from Augsburg, Germany. The tool combines the advantages of several static visualisations, such as tables, line graphs and bar charts into a streamlined display. The application is fully interactive and allows the user to look at the information from any angle she wants, literally.
Effect plots work by identifying high-order terms in a generalised linear model, a statistical technique. Once these terms are identified fitted values are derived and plotted for the relevant groups.
Interactive map and time series (Flu trends)
This map presents an estimate of the intensity of the flu in 20 countries, compared with data for the last six years. The upper panel is a line graph showing the current and past trends of flu intensity by country. The lower panel is the choropleth map with the shading corresponding to the intensity of the flu. The innovative feature of this visualisation is the underlying method of estimation: the intensity of the flu has been approximated based on the number of internet based queries submitted during a certain period of time. Analyses of past data have shown that this method offers good predictions of real levels of the illness.
Visualisation used to show the relative position of research objects. In the example there are four brands rated for several dimensions in a survey.
Scatterplot is used to compare of driving habits and petrol prices. Each point in the plot is joined to the previous years point, with the drawn path indicating order in time.
Tukey's hanging rootogram
This visualisation is a variation of the concept of histograms, combining observed and predicted distributions in a simple way.