Stephen Pearson
Chemical Process Statistician, Syngenta
As Xan Gregg has demonstrated at recent Discovery Summits (https://community.jmp.com/docs/DOC-8270, https://community.jmp.com/docs/DOC-9766), the power and flexibility of Graph Builder makes it a key tool for both exploratory data analysis and presentation. It also means that the number of visualisation possibilities for any given dataset can be overwhelming. So how do you select the most useful graph?
Data visualisation best practise published by people such as Edward Tufte and Stephen Few can help. This talk will demonstrate how the possibilities can be further narrowed with an algorithm that understands the structure of a dataset and uses details about the target audience and the columns that form the basis of the research question.
We will show a scripted application that incorporates thumbnails generated with Graph Builder, along with the process outlined above to present a subset of graphs to a user. Selecting a thumbnail launches Graph Builder with those settings, giving the user a convenient starting point. This can help inexperienced users get started with Graph Builder, avoid common data visualisation mistakes and save time for those with more experience. All of which enables a culture of data-driven decision making.