JMP For:: Analytical Application Development; Business Visualisation; Design of Experiments; Exploratory Data Analysis; Interactive Data Mining; Modeling; Quality Improvement; Reliability; Statistics; Visual Six Sigma

JMP® for Statistics

Through its unique combination of interactive graphics and wealth of built-in statistical functionality, JMP is built for statistical discovery. With software, doing statistics is easy. However, doing useful statistics is not always so simple.

Statistical practitioners who appreciate the close relationship between data and analysis will value the statistical power of JMP and the way its interactivity supports this synergy. JMP provides comprehensive facilities for univariate linear regression, univariate nonlinear regression, the more useful multivariate approaches for exploration and dimensionality reduction, and for time series analysis, among other analysis techniques.

And with JMP, not only can you get your analysis done quickly and correctly, but you can also communicate your findings more easily to other stakeholders.

Nonlinear Fit
Fit Model
Multivariate

JMP’s Nonlinear platform allows you to quickly fit functions that are nonlinear in the effects, using either standard least squares or a custom loss-function. Nonlinear is easy to set up and use, and iterates using either variations of the Gauss-Newton method or the Newton-Raphson method. An extensive pre-supplied function library makes defining your specific analysis easy, and you can quickly add to this library if you need to. Once the fit has converged, you can generate profile confidence limits of the fitted parameters and plot the fitted function. The custom-loss function facility provides additional flexibility, allowing you to use, for example, iteratively reweighted least squares for robust regression.

Here, JMP output shows a series of three fits of mean Response to log concentration using a four parameter Rodbard model with two groups.

The Fit Model platform in JMP provides a unified environment for fitting linear models with specified fixed and random effects and defined error terms. Whatever your favored approach, a complete set of diagnostics and manual and automated fitting approaches allow you to rapidly build useful models. Specific fitting personalities help to focus your attention in the right place, and JMP lets you easily compare the predictive power of competing models, no matter how they were built. Multiple responses are handled in an integrated way, and JMP’s Profiler makes it simple to compare and contrast the interpretability and results of various fits. The Profiler also allows you to find settings that will optimize your response(s), and if required, you can quickly conduct Monte Carlo simulations to assess how variation in the effects will be transmitted into the response(s).

Use JMP for automated fitting of multiple responses from a common base model using validation, with joint profiling.

Multivariate analyses can focus either on observational units (rows) or on variables (columns), and may treat variables on an equal footing or distinguish between effects and responses. But whatever your analytical objective, JMP will work with you to get the job done. In the multivariate context consideration of data quality, the identification and treatment of outliers, and the pattern of missing values are all vital. Typically these issues need to be addressed iteratively as the analysis unfolds, and the interactivity of JMP is built for this way of working. Amongst other analytical techniques, JMP provides: PCA, Hierarchical and K-Means Clustering, Normal Mixtures, Self-Organizing Maps, Discriminant Analysis and PLS. Each Platform uses JMP’s unfolding style of analysis, so that you can shape your approach according to what the data reveals to you.