Application-Specific Help

This guide covers software setup inside the PNR — license activation, paths, and configuration tweaks for tools your team is most likely to use. Self-service installs handle the basics; the per-application sections here cover the gotchas you can't easily discover from a click-to-install dialog.

Overview

This PNR tutorial guides you through:

Installing Applications on a Windows VM
Installing Applications on a Linux VM

Quick Links

NVivo
SAS
Stata
PNR Whisper
Python
VS Code Extensions
R and RStudio Configuration

Working with Your VM in a PNR Project

All Members in a PNR project can build a personal Windows or Linux VM. These VMs can be easily deleted and recreated, allowing you to use the right tools at each phase of your project.

Part 1: Installing Applications on Your Windows VM

Researchers can install software via self-service by clicking the pencil icon next to the Launch Desktop button:

Software Install

Best practice

For best performance, only install the software you need for the current phase of your project. You can delete and recreate your VM later to install other tools.

After checking the boxes for desired software, click Install. You will receive an email confirmation once the installation is complete.

NVivo

NVivo is available in the PNR but requires a license key upon first use and annual renewal.

Visit software.duke.edu
Search for and check out NVivo (free for Duke users)
The license key will be displayed after checkout
Launch NVivo → Enter the license key when prompted

SAS

Setting the Work Library Path

To avoid "insufficient resources" errors caused by temporary files filling your C: drive:

Create a Desktop shortcut for SAS
1. Open the Start menu and search for SAS (for example, SAS 9.4).
2. Right-click the SAS app and select Open file location.
3. In the folder that opens, right-click the SAS icon and choose Show more options → Send to → Desktop (create shortcut).
Edit the shortcut properties
1. On your Desktop, right-click the new SAS shortcut and select Properties.
2. In the Shortcut tab, locate the Target field. It will look similar to:
```
"C:\Program Files\SASHome\SASFoundation\9.4\sas.exe"
```
3. At the end of that line, add a space followed by:
```
-work "P:\read_write\temp"
```
4. The full Target field should now read:
```
"C:\Program Files\SASHome\SASFoundation\9.4\sas.exe" -work "P:\read_write\temp"
```
Apply and test
1. Click Apply, then OK.
2. Double-click the SAS shortcut to launch SAS.
3. In the SAS editor, enter the following command and run it:
```
%put %sysfunc(pathname(work));
```
4. Confirm the log output begins with:
```
P:\read_write\temp\
```

Stata

Stata can be installed on PNR Windows VMs, but it requires a valid license.

Step 1: Purchase a License

To obtain a license, visit software.duke.edu.

Search for and check out Stata.

Step 2: Submit a Help Ticket

Once you have your license details, please submit a Get Help ticket so an RC Admin can apply the license for you.

In your ticket, include the following information:

Licensed software: Stata
License type: (e.g., Single-user, Network, etc.)
License term: (e.g., Annual, Perpetual)
Serial number:
Code:
Authorization:

Note

You must unzip any files before importing them into Stata.

PNR Whisper

PNR Whisper is a Windows desktop application for transcribing audio files in the PNR. It's built using open-source AI models which are known for being highly accurate with automatic speech recognition (ASR) and speech translation. It transcribes audio into text across multiple languages, and has been trained on numerous hours of multilingual, multitask data.

Via the app researchers can select the medium model for fast turnaround or the large model for high accuracy. PNR Whisper supports speaker-labeling so transcripts can better distinguish who is talking. It also includes an automatic translation option to translate non-English audio into English during transcription.

For install just submit a Request Help ticket and pick PNR Project Software Request.

PNR Whisper application window

Dataflow

Open the app on via your desktop
Add individual or multiple audio files
Select medium or large model
Hit start
Transcripts are sent to your project's read_write folder.

Python

Conda Setup

Anaconda Prompt vs. VS Code

You can use the CMD Prompt that comes with the Miniconda install. You can find it in the Start Menu as Anaconda Prompt — this has everything configured automatically.

Alternatively, follow the manual steps below if you prefer to work inside VS Code.

Install Conda and VS Code

In your PNR Windows VM, click the pencil icon to install:
- Conda
- Visual Studio Code (VS Code)
Once installed, open VS Code
- Use the menu to open a new CMD prompt terminal (not PowerShell or Bash).
- Run the following command to activate conda (base): C:\ProgramData\miniconda3\Scripts\activate.bat
Now in the same CMD prompt terminal run: conda init cmd.exe

Note

Disregard the message "Operation Failed" — this is expected. Continue to the next step.
1. Close the CMD prompt terminal. Now when you launch a CMD prompt terminal you will be able to activate conda: conda activate

Set Command Prompt as the default terminal in VS Code

In VS Code press Ctrl + Shift + P.
Type Terminal: Select Default Profile and press Enter.
Select Command Prompt or Command Prompt (Conda) from the list.
Now when you press Ctrl + ` (control and the backtick key) to open a new terminal it defaults to Command Prompt.

Set Conda and Interpreter in VS Code Settings

To make Miniconda the default Python interpreter and automatically activate the base environment in every terminal:

1. Open VS Code Command Palette

Press: Ctrl + Shift + P
Then type: Preferences: Open User Settings (JSON)
Select it.

2. Copy and paste this into the `settings.json` file

{
  "python.defaultInterpreterPath": "C:\\ProgramData\\miniconda3\\python.exe",
  "terminal.integrated.profiles.windows": {
    "Command Prompt (Conda)": {
      "path": "C:\\Windows\\System32\\cmd.exe",
      "icon": "terminal-cmd",
      "args": [
        "/K",
        "C:\\ProgramData\\miniconda3\\Scripts\\activate.bat && conda activate base"
      ]
    }
  },
  "terminal.integrated.defaultProfile.windows": "Command Prompt (Conda)"
}

3. Save and Reload

Save the file: Ctrl + S
Reload VS Code: Ctrl + Shift + P → Reload Window (or close VS Code and reopen)

4. Open the Terminal

Press: Ctrl + `
You should now see the terminal open with: (base) C:\Users\yourname>

Create Conda Environment (example)

Use isolated environments

To avoid dependency issues, best practice is to work in a clean, isolated environment.

Open a new CMD prompt terminal in VS Code
Run the following command to activate conda (base): C:\ProgramData\miniconda3\Scripts\activate.bat
Create the environment: conda create -n torch-env python=3.10
Activate the environment: conda activate torch-env

Tip

You can list all conda environments at any time with: conda env list

Example: Install PyTorch (CPU-Only)

conda config --add channels conda-forge
conda config --add channels pytorch
conda config --add channels nvidia
conda config --set channel_priority strict
conda install pytorch cpuonly -c pytorch

VS Code Extensions

Python/Jupyterlab Extensions

Request extension files

To get Python or JupyterLab .vsix extension files for install in VS Code, please request via ticket after choosing PNR Project Software Request: Request Help.

Once your ticket is complete install your extension .vsix file:

Open VS Code
Press Ctrl+Shift+P
Run Extensions: Install from VSIX
Select the .vsix file (e.g., P:\transfer\to_project\ms-python.python-<version>.vsix)

Or via terminal:

From the folder where you placed the .vsix files, run in a terminal:

code --install-extension ms-python.python-<version>.vsix
code --install-extension ms-toolsai.jupyter-2024.10.2024100401-win32-x64.vsix

Replace <version> with the filename in your folder.

JupyterLab extension version

For compatibility with VS Code version, you must use JupyterLab extension version:

ms-toolsai.jupyter-2024.10.2024100401-win32-x64.vsix

Setting up JupyterLab in VS Code (after installing the JupyterLab extension)

Create a new conda environment

conda create -n jupyter_test python=3.11 -y
conda activate jupyter_test

Install Jupyter kernel support (conda-only)

conda install ipykernel jupyter_core jupyter_client -y

Register the environment as a kernel

python -m ipykernel install --user --name jupyter_test --display-name "Python (jupyter_test)"

Verify the kernel is registered

jupyter kernelspec list

You should see something like:

Available kernels:
  python3       C:\Users\<netid>\.conda\envs\jupyter_test\share\jupyter\kernels\python3
  jupyter_test  C:\Users\<netid>\AppData\Roaming\jupyter\kernels\jupyter_test

Open the Command Palette: Ctrl + Shift + P
Search for and select: Create: New Jupyter Notebook
Save the notebook as: test.ipynb.
In the top right, click Select Kernel → choose Python (jupyter_test).
Run a cell like:
```
print("Hello, world!")
```
Check for output below the cell. If you see Hello, world! the setup is done.

R and RStudio Configuration

Even with proxy set at install, R needs internal proxy settings:

# Set CRAN repo
myrepo = getOption("repos")
myrepo["CRAN"] = "http://cran.r-project.org"
options(repos = myrepo)
rm(myrepo)

# Set Proxy
Sys.setenv("http_proxy"="http://safer-proxy.oit.duke.edu:3128")
Sys.setenv("https_proxy"="http://safer-proxy.oit.duke.edu:3128")

To debug install issues:

options(internet.info = 0)  # Turn on
options(internet.info = 2)  # Turn off

In RStudio

Run the same commands in the console.

Use install.packages("YourPackageName") to install.

Per-session settings

These settings must be re-applied each session.

Part 2: Installing Applications on Your Linux VM

For software installation please submit a ticket: Request Help.

Need Help?

For software not in the self-service catalog, submit a Request Help ticket and choose PNR Project Software Request
Contact rescomputing@duke.edu for help configuring tools, environments, or workflows in your VM