hpcread error
in [SAS]
|
From: David L Cassell on 13 Dec 2006 16:04 psibara(a)gmail.com wrote back to me instead of to the list: > >psibara(a)GMAIL.COM wrote: >> >>Hi SAS Gurus, >> >>Has anyone ever encountered a SAS error of this sort? >> >>This occured whilst a data step merge was executing. >> >>NB: I'm using the SPDE engine... >> >>data X; >>merge W(in=a) Z(in=b); >>by c; >>if b then delete; >>run; >> >>this is the resultant error message: 'error in hpcread 131.' > > > >Could you write back and enclose the relevant 10 or 20 lines >out of your log? I can't figure out what the problem is from >what you wrote. > > >In particular, what you wrote is *not* the form of a SAS >error message, and it is not a standard SAS error citation. >Could one of your files 'W' and 'Z' be coming from an external >database or data source? > > >HTH, >David >-- >David L. Cassell >mathematical statistician >Design Pathways > >Howzit David, > > >Sorry for the late reply... > > >here's the snippet of the SAS log, I do agree with you that this is not a >SAS error as it doesn't get prefixed with 'ERROR:' however this occurs when >SAS is processing the data step in question.... He also attached several hundred lines of log, which appeared to be cut down from several *thousand* lines of log. Here are some highlights: ======================================= 1509 +proc sort data=delete_moves; MPRINT(ITERATE): proc sort data=delete_moves; 1510 + by eqpno actdate act; MPRINT(ITERATE): by eqpno actdate act; 1511 +run; MPRINT(ITERATE): run; NOTE: Sorting was performed by the data source. NOTE: There were 2835 observations read from the data set WORKSPDE.DELETE_MOVES. NOTE: The data set WORKSPDE.DELETE_MOVES has 2835 observations and 17 variables. NOTE: PROCEDURE SORT used (Total process time): real time 0.03 seconds user cpu time 0.01 seconds system cpu time 0.03 seconds . . . . 1513 +%* Remove all the equipment numbers that have been deleted from the data repository; 1514 + 1515 + data dataspde.FLOW_CLASSIFICATIONS; MPRINT(ITERATE): data dataspde.FLOW_CLASSIFICATIONS; 1516 + merge DELETE_MOVES (IN=_1) dataspde.FLOW_CLASSIFICATIONS(IN=_2); MPRINT(ITERATE): merge DELETE_MOVES (IN=_1) dataspde.FLOW_CLASSIFICATIONS(IN=_2); 1517 + BY eqpno actdate act; MPRINT(ITERATE): BY eqpno actdate act; 1518 + IF _1 THEN DELETE;; MPRINT(ITERATE): IF _1 THEN DELETE; MPRINT(ITERATE): ; 1519 + run; MPRINT(ITERATE): run; error in hpcread 131. ERROR: Error on server libname socket. ERROR: . NOTE: The data step has been abnormally terminated. NOTE: The SAS System stopped processing this step because of errors. NOTE: There were 2520 observations read from the data set WORKSPDE.DELETE_MOVES. NOTE: There were 29105165 observations read from the data set DATASPDE.FLOW_CLASSIFICATIONS. WARNING: The data set DATASPDE.FLOW_CLASSIFICATIONS may be incomplete. When this step was stopped there were 29105035 observ ations and 28 variables. NOTE: Compressing data set DATASPDE.FLOW_CLA decreased size by 60.79 percent. WARNING: Data set DATASPDE.FLOW_CLASSIFICATIONS was not replaced because this step was stopped. ======================================= Okay, my best guess is that the problem is the following, cut straight from the SAS webstite's tech support pages: ----------------------------------------------------- When a path specified in a SPDE LIBNAME statement contains double-byte characters, the following errors can occur: ERROR: Some code points did not transcode. ERROR: Unexpected I/O error occurred. ERROR: Error in LIBNAME statement. This behavior occurs when running SAS 9.1.3 Service Pack 4 with a session encoding that is different from the encoding of the double-byte characters used in the path. The errors also occur if the session encoding is UTF-8, as UTF-8 does not support national characters. To avoid the errors, use the appropriate ENCODING= option on a FILENAME statement which points to a file containing the SPDE LIBNAME statement. Then use this fileref in a %INCLUDE statement to execute the LIBNAME statement. Alternatively, do not use double byte characters in the paths for an SPDE library. ----------------------------------------------------- But a significant part of the problem is that you are trying to work with a mammoth data set by sorting and merging, instead of using more efficient approaches. If you only have 2835 observations in the 'delete' data set, build a format or a hash to hold the information, then make a single pass through the larger (now unsorted) data set and toss the 'delete' records. HTH, David -- David L. Cassell mathematical statistician Design Pathways 3115 NW Norwood Pl. Corvallis OR 97330 _________________________________________________________________ Get free, personalized commercial-free online radio with MSN Radio powered by Pandora http://radio.msn.com/?icid=T002MSN03A07001
|
Pages: 1 Prev: Display Month and Year only Next: Error on Object Spawner: |