News on SPSS Statistics Extension Commands

There are two new extensions that can be installed via the Extensions > Extension Hub menu.

Boruta Feature Selection: The STATS BORUTAFEATURES Extension Command

This procedure finds important predictive features (predictive variables) in a dataset for a scale or categorical dependent variable using the Boruta algorithm. It iteratively compares importances of attributes (variables) with importances of shadow attributes created by shuffling case values of the original ones. Attributes that have significantly worse importance than shadow ones are consecutively dropped (Rejected). Attributes that are significantly better than shadows are accepted (Confirmed). Others are classified as Tentative This process is repeated until there are no tentative ones or the iteration limit is reached.

Normality Analysis: The STATS NORMALITY ANALYSIS Extension Command

Well, this one isn't exactly new.  The original version, released Fall 2024, had to be withdrawn, because the author of the underlying R package it uses heavily modified it, and it caused major breakage of the extension command.  (I won't go into the gory details.)  The command is back, mostly, with major changes, so it isn't completely compatible with version 1, and two of the less useful original plots are gone, but on the positive side, the other plots have been reimplemented outside that R module and are much improved.  The actual uni- and multi-variate normality test statistics are unchanged.

For now, NORMALITY ANALYSIS is only available on Windows due to a technical issue on the Mac that causes the install to fail in many cases.  The SPSS team is looking into that.

------------------------------
Jon Peck
Data Scientist
JKP Associates
Santa Fe
------------------------------