Visualizing compositional data on the Lexis surface

Lexis fields

“…Lexis fields offer a new way to summarize multiple Lexis surfaces in a single surface. Other strategies include the composite surfaces of Schöley and Willekens (2017) and the APC curvature plots of Acosta and van Raalte (2019). Small multiples of Lexis surfaces (for example, panel plots of Lexis surfaces), on the other hand, constitute a de-layering (e.g., Remund, Camarda, and Riffe 2018;Kashnitsky and Aburto 2019), as these are spatially disjoint.…”

“…Surfaces are often displayed as heat maps, contour maps, perspective plots, or variants of these things (Vaupel, Gambill, and Yashin 1987). Various kinds of quantities, such as raw magnitudes, differences (Minton et al 2017), excesses (Remund, Camarda, and Riffe 2018;Acosta and van Raalte 2019), ratios (Canudas-Romo and Schoen 2005), intensities, proportions, derivatives (Rau et al 2017), and even compositions (Schöley and Willekens 2017) can be displayed on Lexis surfaces to put age, period, cohort, or other patterns in relief.…”

Geofaceting: Aligning small-multiples for regions in a spatially meaningful way

“…There is no way to efficiently represent in one plot both absolute and relative values. Thus, the first two figures complement each other: Figure 1 uses the stacked bar plot technique to reveal the variation of young adult mortality in Mexican states over time; Figure 2 shows the dominant cause of death with a colored tile plot on a standard Lexis surface, which can be seen as a categorical version of a heatmap (Schöley and Willekens 2017;Rau et al 2018).…”

Lexis Surface Visualisation Workflow

“…Setting the shape of each point to a rectangle of fixed size, the above mapping translates to a cell or 'pixel' of a given colour/shade ( ) whose midpoint is at position {ℎ( ), ( )}. Applications and methodological developments in Lexis surfaces are more common within demography (Schöley & Willekens 2017) than in other fields which work with population data, such as in public health, (Minton et al 2013;Minton et al 2016;Minton et al 2017) despite wide applicability to many forms of substantive population-based problems, and the capacity of such surfaces to allow many thousands of observations to be compared intuitively and effectively. (Vaupel et al 1987) Further discussion on considerations and variations of LS visualisations are presented in the Appendix.…”

“…(Vaupel et al 1987) Despite an additional monograph in 1997 (Vaupel et al 1997) and the use of LS in sundry papers,(e.g. Lindahl-Jacobsen et al 2016;Schöley & Willekens 2017) LS visualization remained relatively little used within, and even less well used beyond, the demographic community. Knowledge and application of LS remained so limited that in 2013 I had unwittingly 'reinvented' the approach as one of two 'new' graphical methods for exploring the components of international mortality trends in a paper in a top epidemiology journal; (Minton et al 2013) only discovering my lack of methodological originality one year later.…”