Computer recommendations

[7maiasmith]

I would like to use SAS University Edition to summarize a large data file (3GB) which is currently in xpt format.  Thus I have to open the file in SAS Universal Viewer; copy-paste it into Excel; save as csv; and read it into SAS.  I then want to summarize a few different variables by subject and by day.  So my question is:

*What kind of computer would be able to handle this? Either a specific make/model, or specifications. I prefer Windows.

*Is there anything else I need to do (e.g. reset the working directory) to make this work?  If so, how do I do it in University Edition?  Most of the official documentation is for regular SAS.

Thank you!!

[Shmuel]

I'm using SAS UE with windows.

As much as I remember there is no limitation of file size.

You better check on google what are SAS UE requirements

  https://support.sas.com/software/products/university-edition/faq/prerequisites.htm

[Reeza]

It depends on what you mean by summarize. If you're using PROC MEANS or a data step you'll be fine. If you need a regression you may have issues but it depends on how much RAM you make available. You can always spin up SAS UE on an AWS machine though and give it a ton of RAM which would ultimately be cheaper. 

Now, some comments on your statements.

---

@7maiasmith wrote:

I would like to use SAS University Edition to summarize a large data file (3GB) which is currently in xpt format.  Thus I have to open the file in SAS Universal Viewer; copy-paste it into Excel; save as csv; and read it into SAS.  I then want to summarize a few different variables by subject and by day. 

Any reason not to read the XPT file directly? PROC CPORT and PROC XPORT should work here. 

---

If so, how do I do it in University Edition?  Most of the official documentation is for regular SAS.

Thank you!!

SAS Studio is SAS 9.4 TS1M6 so most of the documentation applies. You're more limited by the fact that it's in a locked down environment. 

---

*Is there anything else I need to do (e.g. reset the working directory) to make this work?  If so, how do I do it in University Edition?  Most of the official documentation is for regular SAS.

Set a user library to your computer so that space isn't an issue. 

When you do this, SAS does not write temporary tables to the work library, it goes to the user library instead, but this means you need to clean up temporary files yourself. 

You can set your RAM to as high as you need in the VM, I recommend no more than 50% of your computer otherwise it gets two slow. You're limited to 2 cores, but should set it to use both as well. 

---

---

[7maiasmith]

Thanks for the reply.  Do you know if those recommendations work in University Edition?  For some reason the process of reading files into University Edition is very different from that for regular SAS.

[Reeza]

Not really, it's similar to working on a SAS server vs working on a desktop though. The primary difference is making sure you understand to specify the path correctly using Unix notation and realizing it's a mapped drive, not a local drive.

I use SAS UE on my MacBook so that's not the issue here.

[cnm8191]

I use SAS University for graduate study work. Some of the files are pretty large. I have a Microsoft Surface Pro 4. It is an i5 with 8GB of installed RAM and I use Windows 10. I hope this helps!

[Shmuel]

Have you noticed that the computer should be 64 bit ? and

that you need install a virtual machine utility ? as mentioned in the prerequisites?

[cnm8191]

My laptop is 64 bits and I did install the virtual machine utility. I went through the prerequisites when selecting my laptop since I couldn't get SAS University to work on my Macbook.