|
11/19/09
|
37817 - Filter reapplied after editing the original filter condition is not updated.
If a filter has been has been added, saved and applied in a Forecast Studio project, and the results returned by the filter are resorted in a manner other than the default sort order returned by the filter, then any ...
http://support.sas.com/kb/37/817.html,
18KB
|
Problem Note |
|
|
11/5/09
|
37728 - Stratified trend tests in PROC LIFETEST may be incorrect for Peto, Modified Peto and Fleming tests
Stratified trend tests may be incorrect for the Peto, ModPeto and Fleming tests unless the data are sorted by the GROUP= variable.
http://support.sas.com/kb/37/728.html,
19KB
|
Problem Note |
|
|
10/30/09
|
37650 - Variable Clustering node ignores variable with Frequency Role
The Variable Clustering node in SAS® Enterprise Miner will show the Role of each variable when you select the Variables property. One possible Role the node identifies is Frequency. When you run the node, it ...
http://support.sas.com/kb/37/650.html,
20KB
|
Problem Note |
|
|
8/12/09
|
36811 - Incorrect results possible with full-rank parameterization and AGGREGATE option
Incorrect results may be produced by PROC GENMOD if you specify full-rank parameterization (other than PARAM=GLM) for some predictors in the CLASS statement and also use the AGGREGATE option (without specifying any ...
http://support.sas.com/kb/36/811.html,
18KB
|
Problem Note |
|
|
8/3/09
|
36690 - RESTRICT statement may produce incorrect results if TREND= option is specified on MODEL statement
If the TREND= option is specified on the MODEL statement and the RESTRICT statement is used to impose restrictions on the parameters associated with the exogenous variables, then the restrictions are applied to the ...
http://support.sas.com/kb/36/690.html,
20KB
|
Problem Note |
|
|
7/21/09
|
31963 - UCL and LCL values in the PROC GLIMMIX OUTPUT OUT= data set are incorrect when the model contains only R-side random effects
The upper and lower confidence bounds in the OUTPUT OUT= dataset from PROC GLIMMIX are incorrect when the model contains only R-side random effects and the model is not an intercept-only model
http://support.sas.com/kb/31/963.html,
20KB
|
Problem Note |
|
|
7/10/09
|
36068 - Incorrect standard errors when format reorders CLASS levels in PROC SURVEYREG
If a format is specified that changes the internal order of levels of a CLASS variable in PROC SURVEYREG and either METHOD=JK or METHOD=BRR is specified, then the standard errors for estimated regression coefficients ...
http://support.sas.com/kb/36/068.html,
19KB
|
Problem Note |
|
|
12/23/08
|
14760 - Could not create a new MPCONNECT session
After applying Service Pack 2 on a UNIX server, a SAS® Enterprise Miner™ node that uses the UNIX workspace server might result in the following error:
Could not create a new MPCONNECT session.
Please check ...
http://support.sas.com/kb/14/760.html,
19KB
|
Problem Note |
|
|
9/12/07
|
20857 - The OUTPUT data set in PROC PHREG contains incorrect survival estimates for time-dependent covariates
When the OUTPUT data set in PROC PHREG is enabled using the counting
process style of input (MODEL (START, STOP) syntax) for time-dependent
covariates, the survival estimates produced in the OUTPUT dat set are
incorrect ...
http://support.sas.com/kb/20/857.html,
18KB
|
Problem Note |
|
|
3/29/07
|
19736 - Multiple table requests in PROC SURVEYFREQ may result in incorrect table results
When you have multiple table requests in a single PROC SURVEYFREQ step
and there are missing values in any of the analysis variables, the
procedure will report incorrect results for each of the tables.
The only ...
http://support.sas.com/kb/19/736.html,
18KB
|
Usage Note |
|
