06-06-2016 12:55 PM
My dataset has over 500,000 rows of data.
One category field in my data is called cat1, the Model Type is discrete, and the Format is $4.
There are 117 levels (distinct values) for cat1.
Another category field in my data is report_time, the Model Type is Continous, and the Format is Date, Time. There is no duplication in cat1*report_time.
I need to create a filter in SAS VA 7.3 Data Exploration so that I am only pulling data from the record associated with the last submitted date/time.
Then I will create a bar chart of a numeric measure with aggregation of sum.
My goal is to have the bar chart add one value per level of cat1, and that one value will have a variety of report_time. There is no reference time and I can not use a time period for the pull. It must simply be the most recent record for each distinct level of cat1.
The last report_time could be different for each level of cat1 so I can not use the regular boolean operators.
06-06-2016 03:36 PM
I see two possible solutions - both of these are untested.
Hope this gives you an exploration trail.
06-14-2016 10:06 AM
06-14-2016 10:11 AM
Using the TreatAs function you can convert a Date value into a numeric value. Then you could get it's MAX value.