17322 Views 12 Replies Latest reply: Aug 27, 2008 5:46 PM by drannmaria
SAS is not available in a Mac-native version; to the best of my knowledge, of all the major statistics packages, only SPSS is currently available for Macs, though depending on your needs, there may be other solutions available. If your classes require use of SAS, though, it could be run in Windows on a Mac, either via Boot Camp or one of the virtualization solutions such as Parallels Workstation or VMWare Fusion.
I don't know if this will be helpful, but this site has a lot of information on statistical apps for the Mac. Be sure to check the page on software for Intel Macs. There are some excellent packages available that provide most if not all the essential statistical analyses provided by SAS/Stat.
Stata, JMP, and AAbel are major packages. If you don't mind working for it, the open-source R is an SAS/SPSS level statistical package but requires learning its programming quirks. But it does have the advantage of being a major stat package and being free of charge.
I wholly agree. In my last 17 years in academia I used Macs. It was always a challenge to find good statistical software, but there were several decent packages available that ran under Mac OS. But with the advent of the PPC there were fewer providers willing to port their products. One of the best was called StatWorks (I think.) JMP was the SAS Group's sop to the Mac, but I never much cared for it. I don't know if JMP is still made by SAS or if they've spun it off. Today, unfortunately, there isn't a lot of choice if you want a major stat package and Intel support. JMP, Stata, AAbel, and R are just about it. There are some lighter weight packages. I've used InStat in the past which was reasonably decent if you didn't need more than basic stats, hypotheses testing, and regression. But the GraphPad people never ported it or GraphPad to Intel.
If you must use SAS and it doesn't provide a Mac option, don't rule out running it on Apple hardware. I am having great luck using Vmware Fusion to run Windows and a few unfortunate field-specific apps. Of course, there is the BootCamp thing, but I don't know anything about that.
As one who often teaches stats to grad students, I generally discourage the use of "major stats packages." Those who come to me with prior experience in this arena generally know only how to click on menu options or run default scripts and produce large tables of numbers about which they understand very little.
+1 for R
Well one of my first semester classes (I'm in a masters in public health program) is called "Programming in SAS," so I don't think I can avoid it.
I haven't used the a new Intel based Mac yet, but I'm leaning toward doing that and then using Bootcamp as someone suggested. I'm not a Mac expert and I think I would have difficulty working in code, as someone else suggested. I just hope that the MacBooks will be adequate for running the Windows OS and SAS.
Well, I will admit that as a graduate student, I, too, used SAS. As far as major packages go, it does seem to offer the most detailed control of what is going on through its proc statements, and I've worked with competent statistical researchers that used it extensively. Its documentation is voluminous and pretty good. As an academic sysAdmin later in life, though, their licensing was brutal and expensive. I try to focus my work on opensource and free solutions since many of the people I help are in less developed (=poorer) countries - one of my reasons for liking R so much. In my own experience, nothing but pain has ever come from a student coming to me with stacks of SPSS output.
While I don't have direct experience running Windows on a MacBook, my experience with an iMac and Fusion suggest it should do just fine. In fact, I am currently orchestrating the purchase of a MacBook with the intention of setting it up, if necessary, to dual boot Windows for portable access to some heavy-duty 3D laser-scanning software.
If you install Windows via Boot Camp and dual boot, then the MacBook's Windows performance will be the same as a dedicated Windows machine of similar configuration. However, this will not be the case with virtualization software such as Parallels or VM Fusion. Mainly this is because the MacBook's GPU simply isn't as effective with virtualization. On the other hand there shouldn't be much difference in CPU processing power which is what SAS mainly needs. If you decide to use virtualization software then be sure to max out the RAM in the MacBook.
Windows on Intel Macs
There are presently several alternatives for running Windows on Intel Macs.
1. Install the Apple Boot Camp software. Purchase Windows XP w/Service Pak 2 or Vista. Follow instructions in the Boot Camp documentation on installation of Boot Camp, creating Driver CD, and installing Windows. Boot Camp enables you to boot the computer into OS X or Windows.
2. Parallels Desktop for Mac and Windows XP, Vista Business, or Vista Ultimate. Parallels is software virtualization that enables running Windows concurrently with OS X.
3. VM Fusionand Windows XP, Vista Business, or Vista Ultimate. VM Fusion is software virtualization that enables running Windows concurrently with OS X.
4. CrossOver which enables running many Windows applications without having to install Windows. The Windows applications can run concurrently with OS X.
5. Last is Q. Q is a freeware emulator that is compatible with Intel Macs. It is much slower than the virtualization software, Parallels and VM Fusion.
Note that Parallels and VM Fusion can also run other operating systems such as Linux, Unix, OS/2, Solaris, etc. There are performance differences between dual-boot systems and virtualization. The latter tend to be a little slower (not much) and do not provide the video performance of the dual-boot system.
See MacTech.com's Virtualization Benchmarking for comparisons of Boot Camp, Parallels, and VM Fusion.
Boot Camp is only available with Leopard. The Boot Camp Beta that was used with Tiger has expired and is no longer available for use. So contrary to the other poster's comment, Boot Camp isn't truly "free." You must purchase Leopard to get it.
Yes, you can use SAS on a Mac using Virtual PC, Parallels or Vmware. I used Virtual PC for years and it was deathly slow. SAS under Parallels on my MacBook and on a G5 desktop both run as well as on a native Windows machine.
I use it every day with no problem.
I have it installed on Windows XP and I am sitting here now while it installs on Vista under Parallels.