I have a large set of data that contains many users. I need to calculate a percent of total for each user/site combination, but not for the entire data set, replacing an old proc freq. An example would be as follows:
Site | User | Landmark | COUNT | PERCENT |
Arizona | Jake | A | 4 | 25 |
Arizona | Jake | B | 4 | 25 |
Arizona | Jake | C | 4 | 25 |
Arizona | Jake | D | 4 | 25 |
Utah | Tomas | A | 140 | 70 |
Utah | Tomas | B | 30 | 15 |
Utah | Tomas | C | 30 | 15 |
Texas | Juan | A | 2 | 50 |
Texas | Juan | B | 2 | 50 |
Wyoming | Keith | A | 10 | 33.3 |
Wyoming | Keith | B | 10 | 33.3 |
Wyoming | Keith | C | 10 | 33.3 |
My inclination is to use the summary tool, but I'm having trouble configuring it to get the percent column. Any help is appreciated.
Solved! Go to Solution.
Hi @alexg1,
The Summary tool is correct: then join back to your original data on the column used in your GroupBy.
From there the Percent calculation should be pretty straight forward Formula expression.
Hope that helps!
John
Thanks John!