Regarding SPDS Performance

[Bhoopesh]

How to increase the SPDS performance......

By Macro variables....

[Tom]

Did you search the SAS website?  Here is a page with a number of articles:  http://support.sas.com/resources/papers/tnote/tnote_spds.html

[Bhoopesh]

Thanks Tom..Yes will have a look at it...

Do you know how to access two clusters from a cluster table at a time in SPDS using MEMNUM option..

Like for e.g out of 8 clusters..i need to access 4th and 5th cluster..How to do that..

[LinusH]

I don't have SPDS att my fingertips right now, but my interpretation of the documentation is that the MEMNUM option only lets you specify one memnum at a time.

Can you specify what you mean by "access"? If you mean querying, you need to query the cluster twice:

data want;

     set spdslib.mycluster(memnum=3)

          spdslib.mycluster(memnum=5);

...

run;

But why do you need to access it by memnum? Depending on maintenance of the cluster, members can change memnum over time. I think it's more flexible to query based on minmaxvarlist, so yo don't have to be dependant on the internal structure of the cluster.

[Bhoopesh]

Above code is not working...It gives as

ERROR: Invalid MEMNUM value specified for cluster member open.

How to use MINMAXVARLIST on a particular column from a table..Can you send me some pseudo code for it...

Thanks in advance.

[LinusH]

The usage guide have plenty of examples how to create clusters with minmaxvarlist.

In the query, you need no special syntax, just an ordinary where-clause.

The benefit of clusters comes if you are able to split data on one or more classification columns in your data. A typical column cold be a date or other time period value.

The cluster evaluates the where-clause and are then "automatically" reading data from the only necessary member tables.