Prev: ODS options issue
Next: IRRC Governance year
From: Richard on 14 Dec 2009 01:00 Hi A while back I posted an error with genmod under V9.2 (see below) After a bit more investigation I have determined its to do with the amount of missing data in the class variables. If I exclude all the missing data with a where statement it works OK. There is a related SAS note about this here: http://support.sas.com/kb/17/116.html R Original post: Hi I'm getting an error below in GENMOD under V 9.2. ERROR: Write Access Violation In Task [ GENMOD ] Exception occurred at (4166E8E3) Task Traceback Address Frame (DBGHELP API Version 4.0 rev 5) 4166E8E3 07C7E3CC sasgenmo:mcn_main+0x6D8E3 4166B0B1 07C7E3DC sasgenmo:mcn_main+0x6A0B1 4164738B 07C7E768 sasgenmo:mcn_main+0x4638B 416013D5 07C7FF8C sasgenmo:mcn_main+0x3D5 015A27C8 07C7FFA0 sashost:Main+0xC204 0168D038 07C7FFB4 sashost:Main+0xF6A74 7C80B729 07C7FFEC kernel32:GetModuleFileNameA+0x1BA The identical code below runs perfectly in 9.1.3 proc genmod data = long.sam2 descending; class idalias time trans /param=ref; model gpvisits=time trans / dist=bin link=logit type3; repeated subject=IDalias/ ; Removing the CLASS option PARAM=REF seems to fix it. Anyone seen this?
From: Richard on 14 Dec 2009 16:51 On Dec 14, 4:00 pm, Richard <richard.hoc...(a)gmail.com> wrote: > Hi > A while back I posted an error with genmod under V9.2 (see below) > After a bit more investigation I have determined its to do with the > amount of missing data in the class variables. If I exclude all the > missing data with a where statement it works OK. There is a related > SAS note about this here:http://support.sas.com/kb/17/116.html > R > > Original post: > > Hi > I'm getting an error below in GENMOD under V 9.2. > ERROR: Write Access Violation In Task [ GENMOD ] > Exception occurred at (4166E8E3) > Task Traceback > Address Frame (DBGHELP API Version 4.0 rev 5) > 4166E8E3 07C7E3CC sasgenmo:mcn_main+0x6D8E3 > 4166B0B1 07C7E3DC sasgenmo:mcn_main+0x6A0B1 > 4164738B 07C7E768 sasgenmo:mcn_main+0x4638B > 416013D5 07C7FF8C sasgenmo:mcn_main+0x3D5 > 015A27C8 07C7FFA0 sashost:Main+0xC204 > 0168D038 07C7FFB4 sashost:Main+0xF6A74 > 7C80B729 07C7FFEC kernel32:GetModuleFileNameA+0x1BA > > The identical code below runs perfectly in 9.1.3 > proc genmod data = long.sam2 descending; > class idalias time trans /param=ref; > model gpvisits=time trans / dist=bin link=logit type3; > repeated subject=IDalias/ ; > > Removing the CLASS option PARAM=REF seems to fix it. > > Anyone seen this? Further info: It appears that only the first obs has to have missing values for the class variables for this error to occur. (thanks to Srinivas for this info). R
From: Richard on 14 Dec 2009 16:54 On Dec 15, 7:51 am, Richard <richard.hoc...(a)gmail.com> wrote: > On Dec 14, 4:00 pm, Richard <richard.hoc...(a)gmail.com> wrote: > > > > > Hi > > A while back I posted an error with genmod under V9.2 (see below) > > After a bit more investigation I have determined its to do with the > > amount of missing data in the class variables. If I exclude all the > > missing data with a where statement it works OK. There is a related > > SAS note about this here:http://support.sas.com/kb/17/116.html > > R > > > Original post: > > > Hi > > I'm getting an error below in GENMOD under V 9.2. > > ERROR: Write Access Violation In Task [ GENMOD ] > > Exception occurred at (4166E8E3) > > Task Traceback > > Address Frame (DBGHELP API Version 4.0 rev 5) > > 4166E8E3 07C7E3CC sasgenmo:mcn_main+0x6D8E3 > > 4166B0B1 07C7E3DC sasgenmo:mcn_main+0x6A0B1 > > 4164738B 07C7E768 sasgenmo:mcn_main+0x4638B > > 416013D5 07C7FF8C sasgenmo:mcn_main+0x3D5 > > 015A27C8 07C7FFA0 sashost:Main+0xC204 > > 0168D038 07C7FFB4 sashost:Main+0xF6A74 > > 7C80B729 07C7FFEC kernel32:GetModuleFileNameA+0x1BA > > > The identical code below runs perfectly in 9.1.3 > > proc genmod data = long.sam2 descending; > > class idalias time trans /param=ref; > > model gpvisits=time trans / dist=bin link=logit type3; > > repeated subject=IDalias/ ; > > > Removing the CLASS option PARAM=REF seems to fix it. > > > Anyone seen this? > > Further info: > It appears that only the first obs has to have missing values for the > class variables for this error to occur. (thanks to Srinivas for this > info). > R BTW I'm using 9.2 TS2M0 R
|
Pages: 1 Prev: ODS options issue Next: IRRC Governance year |