An Idea Exchange for SAS software and services

by Super User
on ‎11-21-2016 04:03 PM

Which further aggregations would you like to do using an existing aggregation?



by Occasional Contributor CameronF
on ‎11-21-2016 05:14 PM

For instance, I recently tried to emulate proc surveymeans' standard error calculation (VA's SE operation uses proc means). To do this, you need to be able to find the mean of the column, subtract this mean from each observation, and then sum those differences.


The mean for a column can be created as a new aggregated measure, but when then trying to use it in the second aggregation (sum of observations - mean), VA returns an error along the lines of 'unexpected variable type - expecting a measure rather than an aggregated measure'. I don't see why this should be. If I calculate the averages in base SAS prior to uploading the data to LASR, it works fine, but this isn't a viable solution when there are many possible variable combinations. Plus it defeats the point of using VA for on-the-fly calculations.

by Frequent Contributor
on ‎03-01-2017 09:14 AM

This feature would give VA a capability that it's major competitors all seem to have already.


At the moment I can't find a way to something simple like calculate the percentiles for the sum of groups. i.e. the percentiles for total sales by store.



Idea Statuses
Top Liked Authors