Abstract The use of multielement analytical techniques with ever-increasing power of discrimination and extremely low detection limits generates big and trustful geochemical datasets. To impress a geological meaning to these data, exploratory data analysis must follow an approach of gradual increase of complexity, beginning with simple techniques and then following to more high-end ones. Correlation diagram is a graphical method that couple groups of associated variables from a correlation matrix. Since its first proposal in 1965 by Krumbein and Graybill in the book “An Introduction to Statistical Model in Geology” this diagram remained almost forgotten in the literature, possibly due to wearisome and time consuming manual plotting procedures. We present in this paper the EzCorrGraph app, written in HTML5 and JavaScript language, which generates the correlation diagram from a correlation matrix or a list of pairs with a simple and straightforward approach. To validate the results and routines, we compared diagrams generated by the app with some existing diagrams in the literature. The app is freely available on https://ezcorrgraph.firebaseapp.com/ .

中文翻译：

---

相关图：使用 EzCorrGraph 应用程序对地球化学关联进行图形可视化

摘要 使用具有不断增强的辨别能力和极低检测限的多元素分析技术生成了大而可靠的地球化学数据集。为了给这些数据留下地质意义，勘探数据分析必须遵循复杂性逐渐增加的方法，从简单的技术开始，然后再到更高端的技术。相关图是一种图形方法，它从相关矩阵中耦合相关变量组。自从 Krumbein 和 Graybill 在 1965 年在“地质学统计模型简介”一书中首次提出该图以来，该图在文献中几乎被遗忘了，这可能是由于手动绘图过程繁琐耗时。我们在本文中介绍了用 HTML5 和 JavaScript 语言编写的 EzCorrGraph 应用程序，它使用简单直接的方法从相关矩阵或对列表生成相关图。为了验证结果和例程，我们将应用程序生成的图表与文献中的一些现有图表进行了比较。该应用程序可在 https://ezcorrgraph.firebaseapp.com/ 上免费获得。