Using density-distribution sunflower plots to explore bivariate relationships in dense data

摘要: Density-distribution sunflower plots are used to display high-density bivariate data. They useful for data where a conventional scatterplot is difficult read due overstriking of the plot symbol. The x-y plane subdivided into lattice small, regular, hexagonal bins. These bins divided low-, medium-, and groups. In low-density bins, individual observations plotted as in scatterplot. Medium- contain light dark sunflowers, respectively. sunflower, each petal represents one observation. specific number observations. user can control sizes colors sunflowers. By selecting appropriate be obtained that give both overall sense data-density distribution, well points any given region. Sunflower also contrasted with contour kernel- density estimates. appearance these markedly affected by choice smoothing parameters spacing at which probability function evaluated. helpful guiding se- lection distinguishing between chance systematic variation distribution