Please be more specific. What data source? What version of EGuide?
I have noticed that SAS datasets come into the data view in 'chunks'. I suspect that is both for performance (to view the beginning of the data quickly) and memory utilization (EGuide puts data into memory whereas SAS processes one record at a time).
sure thing. I'm running eg 4.2 and these are native sas data sets on our unix server. sometimes, for no reason, sas decides to limit the number of observations to view. what's even more frustrating is that right clicking the data set to view properties -> rows shows the filtered count. I have to slog thru the log to find out the real number of rows
The reason is that EG cannot determine the number of records that would be yielded by a VIEW. The information is not part of the VIEW definition (as it is for SAS data sets), and it could be quite expensive to count, since it requires a query to process it.
So the behavior you see is by design. Note that it affects just the data viewer and properties windows. When you run a task or query, the SAS session doing the work has access to the full set of records that are contained in the VIEW.
Do you know whether SAS is accessing the data via the META libname engine (instead of using the BASE or V9 engine directly)? That might have an effect on how EG interprets the style of access for this data source.
If you access the data via the META engine, that's a layer of indirection that sort of "hides" the location and data mechanism from your SAS process, and thus from EG.