選擇合適的圖表類型往往是製圖表一開始很重要的事情之一。Claus O. Wilke 在《Fundamentals of Data Visualization》的第 5 章,提供了一個快速的「圖表地圖」,根據資料特性找到合適的圖表,這些內容是針對常見圖表用途來分類,是當我們製圖一開始不確定該製作什麼圖形時,一個很好的導覽方式。對於作者建議的圖形內容可以做為參考之一,最後的決策還是要以資料實際呈現需要(例如:資料類內容以及受眾等)作為最終的考量。
總覽不同圖表的應用情境。這系列文章會陸續介紹許多不同類型由 ggplot2 畫出的圖形,不必只依賴慣用的長條圖或折線圖,還能根據資料型態與目的,選擇更合適的呈現方式。
Chapter 5 of Claus O. Wilke’s Fundamentals of Data Visualization provides a comprehensive overview of chart types categorized by their purpose. It emphasizes that choosing the right visualization is often the most crucial step in data communication. The chapter introduces six major categories: amounts, distributions, proportions, x–y relationships, geospatial data, and uncertainty. Amounts are typically shown with bar charts, dot plots, grouped bars, or heatmaps. Distributions can be visualized with histograms, density plots, boxplots, violin plots, or ridgeline plots, depending on the complexity and focus of analysis. Proportions are represented by pie charts, stacked bars, mosaic plots, treemaps, or parallel sets, suitable for varying levels of categorical depth. X–y relationships are best conveyed using scatterplots, bubble charts, slope graphs, correlograms, or line graphs for temporal patterns. Geospatial data relies on maps, choropleths, or cartograms to highlight regional differences. Uncertainty is communicated through error bars, graded confidence intervals, distribution-based plots such as violins or quantile dot plots, and confidence bands for trend lines. This structured taxonomy serves both as a reference and a source of inspiration, helping practitioners move beyond conventional charts and select visualizations that align with the data characteristics and audience needs.