Togaware DATA MINING
Desktop Survival Guide
by Graham Williams
Google

Other Options

Many options are available in R that are not exposed through Rattle.

The variable Rarg[]surrogatestyle sets how the selection process should select the best variable. If set to 1 variables with a large number of missing values are essentially penalised. Defaults to 0.

Include Missing This will set usesurrogate to 2 (the default). The behaviour here is to try each of the surrogates whenever the main variable has a missing value, but if all surrogates are also missing, then follow the path with the majority of cases.

???? This will set usesurrogate to 1. The behaviour here is to try each of the surrogates whenever the main variable has a missing value, but if all surrogates are also missing, then go no further.

Ignore Missing This will set usesurrogate to 0. The observation is not used any further in the tree building.

Max Surrogates: How many surrogate variables to keep at each node (default is 5)

Surrogate Style: Whether the surrogates are sorted by the correct number of classifications or by the percentage correct over the non-missing values.

Copyright © 2004-2010 Togaware Pty Ltd
Support further development through the purchase of the PDF version of the book.
The PDF version is a formatted comprehensive draft book (with over 800 pages).
Brought to you by Togaware. This page generated: Sunday, 22 August 2010