Problem Note 35543: Incorrect Score code when Special Codes values are in the range of a bin value in the Interactive Grouping node
In SAS® Enterprise Miner™, the Interactive Grouping node supports mixed-type data and special user defined coded holdout values. You determine these special codes, or holdout values, based on the business situation. These codes are included in a SAS data set, which SAS calls a Special Codes data set. Each row in the Special Codes data set contains a special code or holdout value that maps to a special missing value. After the Special Codes data set is created, you implement the mapping by setting the Use Special Codes property to Yes, and, by selecting the data set using the Special Codes Data Set property.
If you use a Special Codes data set that includes an interval value which is less than or equal to the value of the Number of Bins property, dupliclate bins in the Interactive window will result. As a result, the score code and the associated labels will be incorrect.
For example, if the Number of Bins property is set to 20, any value from 0-20 in a Special Codes data set will result in that interval value being used in duplicate bins, and very likely in different groups as well. The scorecard generated score code will have an appearance that is similar to the following:
*------------------------------------------------------------*;
* Variable: input;
*------------------------------------------------------------*;
if MISSING(input) then do;
SCORECARD_POINTS = SCORECARD_POINTS + 192;
SCR_input = 192;
end;
else if NOT MISSING(input) AND input < 28 then do;
SCORECARD_POINTS = SCORECARD_POINTS + 162;
SCR_input = 162;
end;
else if NOT MISSING(input) AND input < 39 then do;
SCORECARD_POINTS = SCORECARD_POINTS + 181;
SCR_input = 181;
end;
...
The line "else if NOT MISSING(input) AND input < 39 then do;" is missing a lower bound on the input variable comparison.
The workaround is to set the Number of Bins property value to a value less than all values in the special codes data set.
Select the Hot Fix tab in this note to access the hot fix for this issue.
Operating System and Release Information
SAS System | SAS Enterprise Miner | Microsoft® Windows® for 64-Bit Itanium-based Systems | 5.3 | 6.1 | 9.1 TS1M3 SP4 | 9.2 TS2M0 |
Microsoft Windows Server 2003 Datacenter 64-bit Edition | 5.3 | 6.1 | 9.1 TS1M3 SP4 | 9.2 TS2M0 |
Microsoft Windows Server 2003 Enterprise 64-bit Edition | 5.3 | 6.1 | 9.1 TS1M3 SP4 | 9.2 TS2M0 |
Microsoft Windows XP 64-bit Edition | 5.3 | 6.1 | 9.1 TS1M3 SP4 | 9.2 TS2M0 |
Microsoft Windows 2000 Advanced Server | 5.3 | | 9.1 TS1M3 SP4 | |
Microsoft Windows 2000 Datacenter Server | 5.3 | | 9.1 TS1M3 SP4 | |
Microsoft Windows 2000 Server | 5.3 | | 9.1 TS1M3 SP4 | |
Microsoft Windows 2000 Professional | 5.3 | | 9.1 TS1M3 SP4 | |
Microsoft Windows NT Workstation | 5.3 | | 9.1 TS1M3 SP4 | |
Microsoft Windows Server 2003 Datacenter Edition | 5.3 | 6.1 | 9.1 TS1M3 SP4 | 9.2 TS2M0 |
Microsoft Windows Server 2003 Enterprise Edition | 5.3 | 6.1 | 9.1 TS1M3 SP4 | 9.2 TS2M0 |
Microsoft Windows Server 2003 Standard Edition | 5.3 | 6.1 | 9.1 TS1M3 SP4 | 9.2 TS2M0 |
Microsoft Windows XP Professional | 5.3 | 6.1 | 9.1 TS1M3 SP4 | 9.2 TS2M0 |
64-bit Enabled AIX | 5.3 | 6.1 | 9.1 TS1M3 SP4 | 9.2 TS2M0 |
64-bit Enabled HP-UX | 5.3 | 6.1 | 9.1 TS1M3 SP4 | 9.2 TS2M0 |
64-bit Enabled Solaris | 5.3 | 6.1 | 9.1 TS1M3 SP4 | 9.2 TS2M0 |
HP-UX IPF | 5.3 | 6.1 | 9.1 TS1M3 SP4 | 9.2 TS2M0 |
Linux | 5.3 | 6.1 | 9.1 TS1M3 SP4 | 9.2 TS2M0 |
Linux on Itanium | 5.3 | 6.1 | 9.1 TS1M3 SP4 | 9.2 TS2M0 |
Solaris for x64 | 5.3 | 6.1 | 9.1 TS1M3 SP4 | 9.2 TS2M0 |
Tru64 UNIX | 5.3 | 6.1 | 9.1 TS1M3 SP4 | 9.2 TS2M0 |
*
For software releases that are not yet generally available, the Fixed
Release is the software release in which the problem is planned to be
fixed.
When there are interval values used as part of a Special Codes data set, incorrect score code might be created.
Type: | Problem Note |
Priority: | high |
Topic: | Analytics ==> Data Mining
|
Date Modified: | 2009-04-20 12:58:37 |
Date Created: | 2009-04-13 17:04:43 |