Dashboard building with JMP®
Dashboards provide quick views on the key information scientists and engineers need to make informed decisions and to convey findings to others.
Dashboards in JMP are more than information visualization. JMP provides a complete workflow for quickly building dashboards encapsulating data access, cleanup, model-building, data visualization, organization and sharing needs. By using JMP add-ins, you can share queries, dashboards and visualizations with others. And if you need to share analyses with those who don't have JMP, interactive HTML has selection data filters for more interactivity within your dashboards.
When you have direct access to data in a structured SAS or ODBC-accessed database, JMP lets you quickly set up queries with multi-table joins, filters and table previews so you can get exactly the right subset to explore. Queries can be saved directly for reuse, updated from your imported data table, and shared. With the ability to automatically add new columns of summary data and run post-query scripts to create meaningful graphs, you can move quickly from raw data to insight.
If you don’t have direct access to your data with an ODBC driver, JMP can still access your data. JMP can import flat files, Excel files using a flexible import wizard, and many other file formats easily. Once in JMP, you can take advantage of the query, filtering, summarization and sorting tools afforded by Query Builder. After you create the query, it can be saved as a script and rerun automatically. All you have to do is pull the most recent data, rejoin, and you are ready to move forward.
JMP can also get data from web sources using the JSL HTTP Request function, which allows you to write custom scripts to communicate with external web servers through REST APIs. By using HTTP Request and the JSON parsing function, you can automate the process of fetching data from the web and placing it into JMP tables.
Data Shaping and Pre-Processing
JMP eliminates unnecessary steps in accessing and shaping data for analysis, letting you avoid using other tools to extract data, prepare it and then analyze it. From a single interface, you can access data, join tables, sample, transform variables and create new ones, explore missing values and determine the best way to treat them – easily recoding and reshaping your data as needed. JMP is self-service everything.
Databases are typically structured for efficient data storage and are not analysis-ready. Due to the high velocity and volume of data, it makes sense to store it as efficiently as possible in a database. But to answer broader questions, clean up your data, and discover interesting patterns and trends, you need to do more to the data. One of the easiest ways to create calculated columns in a single click is to create a New Formula Column to access shortcuts to common transformations and summaries. This shortcut menu offers quick access to the Group By functionality, which lets you identify a column of categories (e.g., Day) to use as groups when you create a new formula column on your measure.
When exploring your data, you need to quickly create and customize a variety of multi-element graphs to identify interesting patterns and potential outliers. The Graph Builder in JMP provides an extensive palette of graph types, an intuitive drag-and-drop interface, and the option to save a script to quickly reproduce any graph you create when you retrieve new data. Graph Builder visualizations can serve as individual building blocks of a more complex dashboard. And if you want to change the type of visualization at any point, even if it is in a dashboard, you can without having to recompile or publish anything and without bothering your IT department.
You can also use filters to slice and dice an individual graph or series of graphs by one or more dimensions. This lets you explore how a behavior changes by time, location or another variable. Fortunately, you don’t need to create a custom table and graph for each interesting subset. JMP provides traditional list-based filters, but you can also choose to create a summary graph as a selection filter for another graph. Clicking on your filter graph immediately subsets your data to show a category or time frame of interest.
At some point in your analysis process, you will want to separate data into signal and noise. JMP includes a full suite of linear and non-linear modeling tools to build useful models of your data and also help with model validity through cross-validation techniques.
And JMP includes a full suite of modeling algorithms, screening techniques and clustering tools. No need to utilize open-source software when you want to employ a statistical model in your dashboard.
Dashboards serve one of two purposes: 1) conveying key findings or solutions to decision makers concisely or 2) providing a heads-up display to show patterns in your data in a consistent way. JMP makes building dashboards easy. Simply create a new dashboard, pick a template and then drag and drop reports and tables on to the canvas. This dashboard can be saved as a script so that it can be run any time new data is available, saved as an interactive HTML report to be shared with others that don’t have or have access to JMP or saved as a web report that presents multiple dashboards. You can even embed a data movie of your dashboard in a PowerPoint presentation using animated .gif functionality.
Sharing Results with Others
When you construct a detailed workflow that gives you the results you need, you can share your steps with other JMP users. You may want to simply provide a script that reproduces key data manipulation steps and just the right graph. As you become more comfortable with scripting, you may develop a custom menu and add-ins that can walk a colleague through your workflow.
You can use the rich JMP Scripting Language (JSL) to customize, re-package, or extend JMP functionality for yourself and others. JMP allows you to automate your favorite analyses and customize reports via JSL. Its flexibility permits you to create sophisticated user interfaces, work with other analytical software like SAS, R, Python and MATLAB, and build JMP add-ins – customized statistical discovery applications to share with other JMP users. This enables you to easily spread the use of analytics throughout your organization, wherever data is available and data-driven decisions are needed. Find an add-in or contribute one of your own in the JMP File Exchange.