The SAS Output Delivery System and reporting techniques

Summary reports/graphs: SAS vs Excel, which is easier and better?

Posts: 59

Summary reports/graphs: SAS vs Excel, which is easier and better?

Hi folks,

In my working I have always come across Excel being used to present summary reports such as graphs and charts using any SAS output raw file. Is it easily possible in SAS like that we do in excel and if so does it look as colorful and neat like the excel stuff.

I am sure SAS EG has enough reporting styles but it makes me wonder why most professional companies still rely on excel. Couldnt SAS perform independently all kinds of tasks? More often than not the SAS output dataset or resulting dataset is exported to an excel file in analytics excercise and all the bits and pieces of reporting takes place in excel.

My Boss says use SAS only when the data is huge otherwise stick to excel.

Please explain your comments with examples and appropriate solutions. Thanks


Super User
Posts: 11,107

Re: Summary reports/graphs: SAS vs Excel, which is easier and better?

For a lot of Excel graph users  1) that's the only tool known, 2) Never deal with data sets with millions of records or thousands of variables, 3) willing to spend hours pointing and clicking to make versions of the same graph for different data.

I learned SAS Graph years before I ever (was forced to) graph in Excel. I admit that for some simple graphs it might take me a bit longer to create the first graph but when I'm producing literally hundreds of graphs of different subsets of the data (the SAS BY group processing feature) I can't imagine how the Excel users stand it. And imagine the headaches after preparing some set of charts and then the boss says he wants this bar/ line/ symbol a different color/ fill/ size in a bunch of charts. In SAS I change a couple of parameters in an AXIS, SYMBOL or PATTERN statement and rerun the script.

For additional example of graphs go to and see how many you can replicate in Excel.

And generally using the SAS exported data to "Analyze" in Excel is what I think of a job security or micromanaging or again, just don't know better. For example in the mid 90s I was working with a graduate student who was familiar with Excel but didn't want to learn to use SAS to analyze the data for his project. He did export the data and bring into Excel, ignoring the bit that he lost about 100,000 records due to the 64K record limit in Excel at the time. Then tried to do some summary statistics such as max, min , mean and standard deviation on the remaining data. At 4 hours he still hadn't gotten the results and asked what was going on. I showed him PROC MEANS and accomplished the summary he needed in about 20 seconds, including typing the code.

Ask a Question
Discussion stats
  • 1 reply
  • 2 in conversation