06-14-2018 05:50 PM
I am trying to use surveyimpute on a data set with about 22 variables. About 60% of the observations have complete responses. Missingness is not completely at random and is associated with key demographics. The dataset has a final sample weight as well as 50 replicate weights. I am trying to using surveyimpute to deal with missingness but keep getting the error message below. Any suggestions to resolve this issue are welcome. Thanks.
ERROR: An exception has been encountered.
Please contact technical support and provide them with the following traceback information:
The SAS task name is [SURVEYIMPUTE]
ERROR: Read Access Violation SURVEYIMPUTE
Exception occurred at (23803EC0)
Address Frame (DBGHELP API Version 4.0 rev 5)
0000000023803EC0 000000002834F380 sasqsimp:tkvercn1+0x12E80
00000000237F130A 000000002834FBF0 sasqsimp:tkvercn1+0x2CA
00000000034389DB 000000002834FBF8 sashost:Main+0x10EBB
000000000343E62D 000000002834FF50 sashost:Main+0x16B0D
00007FFD1B2C8364 000000002834FF58 KERNEL32:BaseThreadInitThunk+0x14
00007FFD1B5170B1 000000002834FF88 ntdll:RtlUserThreadStart+0x21
06-14-2018 06:05 PM
You should show the entire code you submitted that causes the error if you want help determining "why". No guarantee but you might get some help.
I have seen such errors arise when using combinations of options that are problematic in combination but not actual syntax errors the compiler can catch.
Sometimes you can start with a very basic run of the proc with your data and add one statement/option until the error occurs to diagnose which specific element when added is at issue. That might give you hint as to either a data issue or let you look closer at the documentation to see why it broke. Or contact Tech support with a reproducible problem.
06-14-2018 06:45 PM
Here is the code
proc surveyimpute data = hints.hints5c1_working2 method=fefi;
var female_wm age_3cat_wm race_4cat_wm income_4cat_wm education_4cat_wm occupation_2cat_wm MaritalStatus_3cat_wm BornInUSA_wm speak_wm HealthInsurance_wm RegularProvider_wm
censreg RUC2013_3cat ;
CLASS female_wm age_3cat_wm race_4cat_wm income_4cat_wm education_4cat_wm occupation_2cat_wm MaritalStatus_3cat_wm BornInUSA_wm speak_wm HealthInsurance_wm RegularProvider_wm
censreg RUC2013_3cat ;
06-14-2018 06:40 PM
If you get that error with the latest release of SAS/STAT (14.3) and within a fresh SAS session, you should definitely get in touch with SAS technical support. They are very responsive. Helping them correct bugs serves the whole SAS Community.