Installation Guide - ESS 412/512 Introduction to Seismology

Installation Guide - ESS 412/512 Introduction to Seismology#

This guide provides setup instructions for the computational labs in this course. We support two package managers: Conda (recommended for most students) and Pixi (faster alternative for advanced users).

Prerequisites#

  • Basic command line familiarity

  • ~2 GB free disk space for environment

  • Internet connection for downloading packages and seismic data

Option 2: Pixi Setup (Advanced/Optional)#

Pixi is a fast, modern package manager that creates reproducible environments. It’s particularly useful for instructors and graduate students working on multiple projects.

Step 1: Install Pixi#

# macOS and Linux
curl -fsSL https://pixi.sh/install.sh | bash

# Windows (PowerShell)
iwr -useb https://pixi.sh/install.ps1 | iex

Restart your terminal after installation.

Step 2: Setup Course Environment#

# Clone the repository
git clone https://github.com/uw-geophysics-edu/ess-412-512-intro2seismology.git
cd ess-412-512-intro2seismology

# Install dependencies (creates pixi.lock automatically)
pixi install

# For graduate students (ESS 512) with additional packages:
pixi install --environment graduate

Step 3: Launch Jupyter#

# Undergraduate environment (ESS 412)
pixi run lab

# Graduate environment (ESS 512)
pixi run --environment graduate lab

Advantages of Pixi#

  • Speed: 3-5x faster than conda for installation

  • Reproducibility: Automatic lock file ensures identical environments

  • No activation needed: pixi run handles environment automatically

  • Cross-platform: Better Windows support

Verifying Installation#

After setup, verify everything works:

# In a Python console or Jupyter notebook:
import numpy as np
import matplotlib.pyplot as plt
import obspy
from obspy.clients.fdsn import Client

print(f"NumPy: {np.__version__}")
print(f"ObsPy: {obspy.__version__}")

# Test data download
client = Client("IRIS")
print("IRIS client connection successful!")

Expected output:

NumPy: 1.x.x
ObsPy: 1.4.x
IRIS client connection successful!

Platform-Specific Notes#

macOS#

Apple Silicon (M1/M2/M3):

  • Both conda and pixi work natively on ARM architecture

  • ObsPy and all dependencies are available for osx-arm64

  • No Rosetta emulation needed

Intel Macs:

  • Fully supported on osx-64 platform

Linux#

  • Ubuntu 20.04+ and similar distributions fully supported

  • May need system packages: build-essential, libgeos-dev

  • If you encounter cartopy installation issues:

    sudo apt-get update
sudo apt-get install libgeos-dev libproj-dev

Windows#

Known Issues:

  • ObsPy installation can be slow with conda on Windows

  • Pixi generally provides faster installation

  • Some FDSN clients may require SSL certificate configuration

Troubleshooting: If you encounter SSL errors when downloading data:

from obspy.clients.fdsn import Client
client = Client("IRIS", force_redirect=True)

For detailed Windows setup, see: https://docs.obspy.org/

Troubleshooting#

Conda environment creation fails#

# Try using mamba (faster solver)
conda install mamba -c conda-forge
mamba env create -f environment.yml

“Command not found: jupyter”#

The environment may not be activated:

conda activate ess412

Import errors for specific packages#

Reinstall the package explicitly:

conda activate ess412
conda install -c conda-forge obspy

Pixi installation slow or failing#

Check your internet connection and try with verbose output:

pixi install -v

If issues persist, fall back to conda setup (Option 1).

Data download timeout errors#

# Increase timeout in ObsPy client
from obspy.clients.fdsn import Client
client = Client("IRIS", timeout=300)  # 5 minutes

Jupyter kernel not found#

# Conda
conda activate ess412
python -m ipykernel install --user --name ess412 --display-name "Python (ESS412)"

# Pixi
pixi run python -m ipykernel install --user --name ess412-pixi

Getting Help#

  1. Check existing issues: uw-geophysics-edu/ess-412-512-intro2seismology#issues

  2. ObsPy documentation: https://docs.obspy.org/

  3. Course Canvas: Post questions in the technical help forum

  4. Office hours: See syllabus for schedule

For Instructors/TAs#

Setting up the development environment#

# With Pixi (recommended for testing)
pixi install --environment dev
pixi run --environment dev lab

# With Conda
conda env create -f environment.yml
conda activate ess412
conda install -c conda-forge black pytest nbval

Running notebook tests#

# Pixi
pixi run test

# Conda
conda activate ess412
pytest --nbval notebooks/

Generating solution notebooks#

Solutions are kept in notebooks/solutions/ (not committed to student-facing repository).

Contents