topic Re: Data preparation option in SAS Visual analytics in SAS Visual Analytics

Data preparation option in SAS Visual analytics

JM_VFAU — Thu, 14 Oct 2021 03:27:16 GMT

Is there a feature of joining two data sources in SAS VA as part of the data preparation step? If this is possible, is there any special consideration that need to be taken into consideration for the CAS server setup? for e.g if we need to join 2 data sets with 2+ years of data is ok to do the data preparation within SAS VA or outside of SAS VA.

Re: Data preparation option in SAS Visual analytics

SASKiwi — Thu, 14 Oct 2021 06:02:53 GMT

I recommend joining outside of SAS VA especially if you want a repeatable, automated process. Typically I do all VA data preparation in SAS code, then run this as a scheduled job.

Re: Data preparation option in SAS Visual analytics

acordes — Thu, 14 Oct 2021 07:44:46 GMT

As @SASKiwi said, it's a better choice to do it in sas studio via code or through a data plan.

Nevertheless it can be done within the VA report. Right-clicking on the data table gives you several options to join tables.

What you need to have in mind is the following:

calculated items are not part of the join
you cannot define a where clause in the join
the report user can only access the resulting table if he has access to the intervening caslibs of the join.
you mentioned 2 years, have in mind that tables can get unloaded from memory. the join to succeed requires the table to be promoted or elsewise having a file with the same name (sashdat or other formats) in the same caslib

Re: Data preparation option in SAS Visual analytics

Madelyn_SAS — Thu, 14 Oct 2021 13:03:14 GMT

It is not recommended to join tables in the Visual Analytics report itself, as it can cause performance problems.

Re: Data preparation option in SAS Visual analytics

JM_VFAU — Thu, 14 Oct 2021 22:21:15 GMT

Thank you @Madelyn_SAS @SASKiwi @acordes. Very helpful tips.

I under SAS VA has limited capability to join data sets. Listing some of options possible to combine data sets

1) Use Data Studio to prepare data by joining various data sources for exploration purpose. This amount of data is limited by the infrastructure resorces.

2) Once the data exploration is done, prepare the data outside of SAS Viya environment if you need the same data set to have 4 or 5 years data.

Does this sound right?