Model Quality Diagnostic Plots Template for Spotfire® using Spotfire® Data Science - Workbench 1.0.0

Overview

This template is computing various diagnostic charts for evaluating and comparing classification model quality (namely ROC, Gains and Lift charts, in addition to that confusion matrices and error rate behavior with changing cut-off points). It is not doing the actual modeling, it is purely for evaluating the performance of the models based on data input with predictions. The form of the input table is expected to be in this shape: response (actual observed value) and probabilities associated with the predicted category in question. There can be one or more probability columns (typically related to different models).

The theory behind various diagnostic charts can be found in this article. Also, this topic was discussed at Dr. Data Science session. Some of the visuals from this template have been used there:

It is the template where you can change the input data and compute outputs without any changes needed in settings and at the same time, it is content to help users in building Statistica data functions.

What features the data function is showing:

• Example of data function using complex data transformation logic 
• Example of data function using more input data sets as well as more output data sets, in addition, parametrization of variable selection is used
• Example of triggering the data function on pressing a button?
• Example of the universal data function (can be applied on different data without changes in any settings and without changes to underlying workflow)

More info about Statistica data functions is here.

If you are interested in building the Statistica data functions, please find additional examples here.