NEWS

Spring Semester will be More of the Same—We're Ready

The Spring semester will be another pandemic semester, so we wanted to remind you how can use SSCC's core services while staying safe.

Our goal is to provide everything we normally provide, so if you need something, just ask!

Winter Updates and New Servers

We've carried out our Winter software updates, so many, many programs have been updated to their latest version. With this update we moved to the R standard of creating a separate package directory for each version. The good news is that you'll never have to worry about running an older package written for an older version of R. The bad news is that you'll need to reinstall the packages you use after each upgrade. Note that RStudio will automatically identify when a script uses packages you have not yet installed and makes it easy to install them.

We've also put into production the four new Linux servers funded by the Economics department and a Research Core Revitalization Grant. These servers have 48 cores, 768GB of RAM, and, in a first for the SSCC, an NVidia T4 GPU. Most SSCC members don't need GPU, but they're very helpful for some machine learning and natural language processing problems, among others. Two of them have been temporarily placed in Silo for COVID-19 work, and are available as linsilogpu001 and linsilogpu002. The others have been placed in the Linstat cluster as linstat5 and linstat6.

Spring Training

Our Spring training starts off early with our core R and Stata classes, and we'll have some topical workshops after the semester starts. This includes new topics like Data Visualizations with ggplot and Choosing a Qualitative Data Analysis Program.

But we also understand the lack of enthusiasm for yet more online classes, and have been putting a lot of effort into supporting independent learning. The complete RStudio Basics and Data Wrangling in R curriculum is now online, and we're offering office hours during the semester for anyone learning R, Stata, or Python on their own.

For details and to register visit our training page.

Spring Semester Instruction

If you are teaching a course this semester and would like your students to be able to access SSCC resources, please email Caitlin at ctefft@ssc.wisc.edu as soon as possible to make arrangements. Students with SSCC accounts can use Winstat to run statistical software and get help from SSCC's statistical consultants with class work.

If you are planning to teach in person and wish to use SSCC's on campus resources it is particularly important you contact us as soon as possible so we can discuss your options for safe access.

MacOS 10.13 now Unsupported

MacOS 10.13 is now unsupported, meaning it will no longer get security updates. Campus policy dictates that computers running any unsupported operating system should not be connected to the campus network, and we are now blocking computers running MacOS 10.13 from connecting to the SSCC network via VPN.

Upgrades to MacOS are free. If you need help upgrading, contact the Help Desk.

Tip: Get the Most Out of Your Email to the Help Desk

With so much support happening by email, we thought we'd give a few tips for what to include in your email so we can help you right away. We don't mind asking for information we need, but every email back and forth slows down the process of getting you the help you need.

  • If you contact us using a non-UW email address, please mention your SSCC username so we know who you are.
  • Tell us what computer or device you are using. The solution to a problem may be very different depending on whether you're using a personally-owned computer, a University-owned/SSCC-managed computer, or a server like Winstat or Linstat.
  • Tell us what you're trying to do and then describe in some detail what goes wrong. "I can't log into Winstat" doesn't help us very much, but something like "When I try to log into Winstat it accepts my username and password but says it can't load my profile" will give us a very good sense of what's wrong. Including the actual error message (either via copy and paste or a screenshot) is best of all.
  • If you've already taken some steps to try to fix the problem, like rebooting your computer, mention that too.

Some additional thoughts for those with statistical questions:

  • If your question is about a particular command, please include in your email the exact command you ran and the output or error message you received.
  • It's even more helpful if you send us your (non-sensitive) data and code so we can look at the data, run the code ourselves, make changes, etc. But don't send us your entire research agenda! Save the required data right before the command in question, and send us that along with a short script/do file that reads the data in properly and then runs the command in question.
  • This is a good chance to practice writing portable code, so you can share with others when you're done. Write your code such that if we put your files in a folder and make that folder the working directory, your script/do file will run without us having to change it. Our Data Wrangling KB articles all discuss how to write portable code.