sciware

Sciware

Summer Sciware 1

Setting up a laptop for scientific computing

https://sciware.flatironinstitute.org/40_SummerIntro

https://github.com/flatironinstitute/sciware/tree/main/40_SummerIntro

Who We Are

• Sciware is a grassroots, scientific software learning community at Flatiron Institute
• We’re here to help! Our goals are:
  • to help you setup and learn the computing tools you need to do your summer work, and
  • to help you build fundamental computing skills (terminal, VS Code, git/GitHub, Python project setup, etc).
• These will be useful in any kind of computing career, academic or not!

Summer Sciware Sessions

• Schedule
  • June 3: Summer Sciware 1 - Setting up a laptop for scientific computing
  • June 12: Summer Sciware 2 - VS Code and GitHub
  • June 24: Summer Sciware 3 - Intro to the cluster
  • July 2: Summer Sciware 4 - Cluster hands-on
  • All 10 AM - noon in the IDA Auditorium (162 Fifth Ave, 2nd Floor)
• We are recording. Link will be posted to https://sciware.flatironinstitute.org/

Where to Find Sciware

• The #code-help channel on Simons Foundation Slack
  • On your laptop, go to simonsfoundation.slack.com, log in, and add the #code-help channel.
  • Put up your green sticky 🟩 if you’ve found #code-help, or your yellow sticky 🟨 if you can’t find it!
• Other Slack channels:
  • The #sciware channel on Simons Foundation Slack
  • The anonymous Summer 2025 Flatiron Computing Question Form Google Form
  • Sciware website: https://sciware.flatironinstitute.org/

Today’s Agenda

• Install prerequisite software (xcode on Mac, WSL on Windows)
• Install uv Python
• Install VS Code
• Setup a Python project
• Setup a Jupyter Notebook in that project

Prerequisite Software

Put up your green sticky 🟩 when you’re done this slide, or your yellow sticky 🟨 if you get stuck!

• Mac users:
  1. Open a terminal: press Command (⌘) + Space to open Spotlight Search, type “Terminal”, and press Enter.
  2. Run: xcode-select --install
• Windows users:
  1. In the Start Menu, search for Window Command Prompt
  2. Right click on Windows Command Prompt and select “Run as administrator”
  3. Run: wsl --install
  4. Restart your computer
  5. If you can open WSL from the Start Menu, you’re done! Otherwise, ask for help.

More detailed instructions here: https://learn.microsoft.com/en-us/windows/wsl/install

Install VS Code

• Go to code.visualstudio.com
• Click Download in the top right
• Mac users: download and install the Mac version
• Windows users: download and install the Windows version
  • Most of the time with WSL, you will install the Linux version of packages. VS Code is the only exception to this rule!

Windows Only: VS Code Setup

• Mac users:
  • Skip this slide.
• Install the WSL extension
  • Open VS Code from the Start Menu
  • Install the WSL extension
    • If you see a pop-up asking if you want to install the WSL extension, click “Install”
    • If you don’t see a pop-up, click the Extensions icon in the left sidebar (or press Ctrl + Shift + X)
    • Search for “WSL” in the Extensions Marketplace
    • Click the “Install” button next to the “WSL” extension
• Connect to WSL
  • Click the blue “Open a Remote Window” button in the bottom left corner of the window
  • Select “Connect to WSL”
  • VS Code should reload, and you should see “WSL” in the bottom left corner of the window

Open a Terminal in VS Code

• Mac users:
  • Open VS Code from the Applications folder
  • Press Command (⌘) + J to open the terminal
• Windows users:
  • Open VS Code from the Start Menu
  • Press Ctrl + J to open the terminal

Install uv Python

Put up your green sticky 🟩 when you’re done this slide, or your yellow sticky 🟨 if you get stuck!

• Go to docs.astral.sh/uv/
• Click Installation on the left
• Copy the standalone installer command: curl -LsSf https://astral.sh/uv/install.sh | sh
• Paste the command in the terminal and press Enter
• Close the terminal (type exit and press Enter)
• Reopen the terminal (Ctrl/⌘ + J) and run uv --version to check that it installed correctly

Start a Python Project

These instructions are adapted from uv’s “Working on projects” documentation.

• WSL Users: check the Remote Window icon in the bottom left of the screen: it should say WSL. If it does not, click the icon and select “Connect to WSL”.
• Open a terminal in VS Code (Ctrl/⌘ + J).
• To create a new Python project and open it as a VS Code workspace, run:

  uv init sciware-2025
  code sciware-2025
  

uv Cheatsheet

| Traditional Command | uv Equivalent | Description | |—————————-|———————–|———————————————| | pip install <package> | uv add <package> | Add a package as a dependency | | pip uninstall <package> | uv remove <package> | Remove a package from dependencies | | python script.py | uv run script.py | Run a Python script using project env | | python | uv run python | Get an interactive Python prompt |

Add a Dependency

• To add a dependency (for example, NumPy), use:

  uv add numpy
  

• Check the pyproject.toml file—uv has added NumPy as a dependency!

Run a Script

• Create a new file (e.g., science.py) and write code that uses your new dependency.
• To run your script, use:

  uv run science.py
  

• With uv, use uv run script.py instead of python script.py to ensure your virtual environment is always in sync with your dependencies.
• After adding a dependency or running a script, a virtual environment will appear in the .venv directory.
• Make sure VS Code uses this environment: press F1, type Python: Select Interpreter, and choose the .venv in your project directory.

For more advanced projects (like packages for GitHub or PyPI), see uv’s project concepts.

Put up a green sticky 🟩 if running a script works, or your yellow sticky 🟨 if you get stuck!

Using Notebooks

These instructions are adapted from uv’s “Using Jupyter from VS Code” documentation.

• Add ipykernel as a development dependency:

  uv add --dev ipykernel
  

• Make sure VS Code is using the virtual environment:
  • Press F1, type Python: Select Interpreter, and select the .venv in your project directory.
• Create a new notebook:
  • Press F1, type Create: New Jupyter Notebook.
  • Click Select kernel in the top right, choose Python environments..., and select the .venv in your project directory.
• To add new dependencies (like NumPy), run uv add numpy in the terminal—they’ll be available in your notebook.

Exercise: Create a New Project

Put up your green sticky 🟩 when you’re done this slide, or your yellow sticky 🟨 if you get stuck!

1. Create a new project called sciware-exercise in your home directory (not in the sciware-2025 project that we already created).
2. Open the project in VS Code.
3. Add numpy and matplotlib as dependencies.
4. Add ipykernel as a development dependency.
5. Create a new notebook in the project.
6. Make a colorful scatter plot from a 2D Gaussian distribution!
   1. Use NumPy to generate 100 random points from a 2D Gaussian distribution with mean [0, 0] and covariance [[1, 0.5], [0.5, 1]].
   2. Use Matplotlib to create a scatter plot of the points, with each point colored by a random value between 0 and 1.

Exercise: Solution

import numpy as np
import matplotlib.pyplot as plt

N = 100
rng = np.random.default_rng()
mean = [0, 0]
cov = [[1, 0.5], [0.5, 1]]  # Covariance matrix for some correlation
x, y = rng.multivariate_normal(mean, cov, N).T
colors = rng.random(N)  # Random color for each point

plt.scatter(x, y, c=colors, cmap='viridis', s=60)
plt.title('Random Colorful Gaussian Scatter Plot')
plt.xlabel('x')
plt.ylabel('y')
plt.colorbar(label='Color value')
plt.show()

Resources

• uv has excellent documentation, including a guide for Jupyter integration. If you get stuck, check the docs, or get in touch with us!
• Microsoft’s VS Code introduction: https://code.visualstudio.com/docs/getstarted/getting-started
• VS Code documentation for the cluster: https://wiki.flatironinstitute.org/SCC/Software/VSCode

Survey

Please give us some feedback!