The computing in this class will use the University of Michigan’s cloud computing system called Greatlakes. To access this system, you will need access to a computer with a web browser, a VPN client (if working off-campus), and a stable internet connection. You will not need to install any other software on your computer.

All students in this course must request a Greatlakes account by filling out this form. Please do this as soon as possible as it may take a few days for the request to be processed. When filling out the form, enter the name of your Statistics 206 instructor as the “advisor”, enter “Statistics 206” under the project description, and enter “Literature, Science, and the Arts” in the college/department field. For the final question on the form, the ARC-TS service to which you are requesting access is “Great Lakes” (do not select any of the other services except for Great Lakes).

You will need to run a VPN client on your computer when accessing Greatlakes from off campus. Information about obtaining and configuring free VPN clients for all major platforms can be found here. If you are working on the UM campus, you do not need to use a VPN.

Greatlakes uses two-factor authentication. You will need to have your phone or other two-factor device with you whenever you plan to use this system.

The IT staff must manually add everyone from the Statistics 206 roster to the account for this course. Near the beginning of the semester, everyone who is currently enrolled in Statistics 206 will be added. If you add the class late and cannot access the system, let your instructor know that you need to be added to the account.

Connecting to Jupyter – step-by-step instructions

The following five steps will create a Jupyter session for you. You will need to follow these five steps every time you want to use Jupyter on Greatlakes.

1. If working off-campus, make sure your VPN client is activated.

2. Connect to Greatlakes using your web browser. If you click this link and get a message saying something like “this site can’t be reached”, then most likely you are off campus and your VPN is not turned on.

3. After connecting to Greatlakes, select the “Interactive Apps” menu at the top of the window, then press “Jupyter Notebook”.

4. You will be taken to a form that allows you to request access to the Jupyter notebook server. Most fields in the form will be set to default values that you do not need to change. Under account, you should enter “stats206s###f21_class”, where ### is your three digit section number, either “001” or “002”. By default, you will be given access for one hour. You can change this to a greater value, but requesting very long sessions may rapidly deplete your $100 computing allocation. You should always request 1 core and 4 GB memory per node (changing these to greater values will not provide any benefit, and may rapidly deplete your allocation).

5. Press the Launch button and you will be taken to another page showing that your request is processing. Within a minute or two, a button that says “Connect to Jupyter” should appear. Press this button and you will be directed to a file listing for your home directory.

Obtaining the notebooks

The first time that you connect to Jupyter, you should perform the following steps to obtain local copies of all the notebooks that you can execute and edit. You will only need to do this one time (if you do it again, it will over-write any changes to the notebooks that you have made).

1. Near the top right corner of the screen, there is a menu that says “New”. Press it, then select “Terminal”.

2. At the prompt in the terminal window, type “cd” then press “enter”. Then type “pwd” followd by “enter” and you should see the path to your home directory, which will have the form “/home/uniqname”, where “uniqname” is your UM uniqname.

3. Type or copy/paste the following command into the terminal and press “enter”:

wget raw.github.com/kshedden/stats206/main/case_studies/download.sh -O download.sh

4. Type “bash download.sh” then press “enter”.

5. To confirm that everything worked, type “ls -lt” followed by “enter”. You should see the names of several files ending with “.ipynb”. These are the Jupyter notebook files that we will use in this course.

You can open any notebook by double clicking the file name in the directory listing. Once the notebook is open, you can run it by clicking on “Cell” in the top menubar, then selecting “Run All”.