Data visualisation and exploration for pharma applications using JMP®

Process understanding provides the only sure way to optimally manage real-world processes, and the pharma industry is no exception to this. Generating such understanding relies on working effectively with the empirical data generated by measurements of process inputs, outputs and ancillary factors to control and improve outcomes while complying with regulatory requirements. The volume, velocity and variety of the data generated in routine operations mean that tabulations of data are not enough, and graphical visualisation is the only way to tame this complexity and make useful progress.

Visualisation and EDA with JMP®

JMP provides all the tools needed to make breakthrough discoveries from data and convey results. By combining dynamic graphics with sophisticated statistics, JMP helps to develop a narrative and interactively share findings in ways decisionmakers can easily comprehend and act on.

Discerning patterns in data and making discoveries is easy in JMP. One can manage and study data while in the flow of analysis and rerun commands as new questions and ideas arise. Furthermore, JMP is desktop software, and it eliminates the need to wait for a server to send findings to a client, especially when dealing with large amount of data.

JMP allows heuristic, open-ended and dynamic Exploratory Data Analysis (EDA), which frequently entails handling data quality and aggregation issues as one examines the data and plays
with different representations to let the data speak clearly. EDA in JMP also helps to develop a useful statistical model when this is required.

Importing and connecting data sources
Data collection and manipulation are the first steps in one’s analytic journey – and frequently, the most time-consuming. EDA helps discover the structure in data, whether in small samples or in the vast amounts of data collected from many processes over time. JMP provides the tools one will need to speed this critical stage of the analysis workflow with a robust range of capabilities for accessing, combining, filtering and cleansing data in preparation for any analysis. JMP’s interactive visuals and extensive analysis capabilities make it an excellent alternative to spreadsheets for EDA and any sort of statistical analysis.

Interactive graphing and analytics at its finest
Thanks to its dynamic linking, in one graph or data table are mirrored in all graphs that are based on that table. The wide repertoire of graph types allows to display and explore the data in any way, helping one to formulate questions that the data may be able to answer.

One can pan and examine JMP’s graphical displays with various tools, zoom in for a closer look and access all required details on demand. JMP’s unique graph builder allows one to create displays through drag and drop, with multiple X and Y grouping variables and a variety of graph elements such as bar charts, histograms, line charts and contour plots. Using these such displays in conjunction with dynamic linking and filtering allows one to work effectively with many variables at once. Graphs can also be used to filter other graphs.

Drag and drop graphing, filtering and column switching

JMP’s drag-and-drop graphing allows one to build and alter graphs quickly. One can make a wide range of charts and plots such as spatial and geographical maps, line graphs, treemaps, waterfall plots and heatmaps to explain one’s findings. Graphs can also be saved as interactive HTML to share with colleagues.

JMP allows one to manage which data to include in graph by using the local data filter, which allows one to interactively define and select complex subsets of data, hide them in plots, or eliminate them from further consideration. The column switcher allows one to easily study different variables without having to redo your analysis. One can record resulting visualisation as an animated.gif and include this dynamic representation in a PowerPoint presentation to share with someone who doesn’t have JMP.

Geospatial maps
JMP provides out-of-the box geospatial mapping at any level of granularity. JMP has prebuilt maps for countries, states and other locations, and allows one to build his/her own without the
need to write code.

JMP also allows one to plot point-oriented data on background maps, showing one’s points in relation to cities, roads and water bodies. This provides additional geospatial context that strengthens one’s narrative, offering new insights that would otherwise be difficult or impossible to uncover. A JMP server hosts the map data used to automatically generate the requisite
OpenStreetMap (OSM) images.

Graphs worthy of sharing and publication
The goal of one’s analysis is not just to provide eye-catching graphs, but to aid in answering one’s questions and formulating interesting new ones. Similar to EDA, visualisations can also be used to display the results of statistical modelling, whether it’s showing the distribution of inputs that affect the outcomes of interest, or graphically illustrating the relationships betweenthese variables using the interactive profiler template-driven drag-and-drop interface, with the option to easily share this with others through an interactive HTML report or JMP Add-In.

Resources
Whitepaper – Effective Visualization Techniques for Data Discovery and Analysis
https://www.jmp.com/en_us/whitepapers/jmp/effective-visualization-techniques.html

Video – JMP Graph Builder Graph Builder (Interactively create visualizations to explore and
describe data).
https://www.jmp.com/en_nl/learning-library/topics/graphical-displays-and-summaries/graphbuilder.html

Download your free trial of JMP®
Used by hundreds of thousands of data explorers worldwide, JMP data analysis software reveals
insights that raw tables of numbers or static graphs tend to hide. Get more out of your data by
downloading a free, fully functional 30-day trial now

Download JMP Trial | JMP