Tweetstorm:
Which is better, R or SAS?
Which is better, R or SAS?
I've heard that R is a bit like a Swiss army knife. It has a package that can do almost anything.
In R's open source environment, packages can be written to apply cutting edge methods as soon as they are available.
SAS is developed in a corporate setting, and you might be waiting a long time before you can try that cool trick you read about online.
SAS is developed in a corporate setting, and you might be waiting a long time before you can try that cool trick you read about online.
But ANYONE can write a package in R. Some are pretty iffy.
In SAS you don't have to worry. The reputation of the corporation depends on its ability to deliver a quality product.
In SAS you don't have to worry. The reputation of the corporation depends on its ability to deliver a quality product.
And what about support? If I've paid out for a SAS license, I sure better be getting help when I need it.
With R, you've got the online user community. Usually crowd sourcing an answer works great. But sometimes there may not be many users out there who can answer your question.
With R, you've got the online user community. Usually crowd sourcing an answer works great. But sometimes there may not be many users out there who can answer your question.
In R there are always a dozen different ways to solve a problem. In SAS there is much more uniformity.
When I write SQL in SAS I am confident that my co-workers will understand my program. With R, everyone writes SQL differently. That's a barrier to collaboration.
When I write SQL in SAS I am confident that my co-workers will understand my program. With R, everyone writes SQL differently. That's a barrier to collaboration.
Another note on service:
If something in R is malfunctioning, it is not clear who is accountable.
With SAS, I know exactly where to direct my anger.
If something in R is malfunctioning, it is not clear who is accountable.
With SAS, I know exactly where to direct my anger.
But if R is a Swiss army knife, then SAS is a CHAINSAW.
It has something that SAS hasn't got: an engine.
It has something that SAS hasn't got: an engine.
SAS can chew through data by running processes on external servers, with parallel processing, and with compression.
It's much more difficult to apply some of these methods in R. I'm limited to the 8G of RAM on my computer, and I'm constantly bumping my head on the ceiling.
It's much more difficult to apply some of these methods in R. I'm limited to the 8G of RAM on my computer, and I'm constantly bumping my head on the ceiling.
The moral? In my case, I've found that SAS is better for work with very large data sets and for standardized processes.
But R has served me better for data exploration and visualization. Plus it's free. Can't argue with that 🤑
But R has served me better for data exploration and visualization. Plus it's free. Can't argue with that 🤑
What are some of the pros and cons you've found about working with R, SAS, or another stats software?
These thoughts courtesy of my friends in the Data Analysis Club at StatCan
(PS - my tweets should never be taken as opinions of StatCan)
(PS - my tweets should never be taken as opinions of StatCan)
