Paths in Python on DesignSafe

Paths in Python on DesignSafe#

Writing Portable, Environment-Aware Code

When working in Python—especially across JupyterHub, HPC jobs, and Tapis-launched workflows—hard-coded strings quickly become brittle. Python provides tools that let you compose, resolve, and validate paths safely, regardless of where your code runs.

Why pathlib Matters#

Python’s pathlib module treats paths as objects, not strings. This makes code:

  • More readable

  • Less error-prone

  • Easier to adapt across environments

from pathlib import Path

base = Path("/home/jupyter/MyData")
input_file = base / "models" / "frame01.tcl"

print(input_file)
# /home/jupyter/MyData/models/frame01.tcl

Advantages over string paths:

  • Automatically inserts /

  • Supports .exists(), .is_dir(), .resolve()

  • Works naturally with relative paths

Relative vs Absolute Paths in Practice#

In notebooks, relative paths often work because the current working directory (CWD) is stable:

Path("inputs/model.tcl")

But in batch jobs or Tapis apps, the CWD may be:

  • A temporary execution directory

  • A staging folder created at runtime

Best practice: Resolve paths early and explicitly.

model_path = Path("/work2/05072/silvia/stampede3/projectA/input/model.tcl")

Using Environment Variables#

DesignSafe and TACC define useful environment variables:

import os
from pathlib import Path

work = Path(os.environ["WORK"])
scratch = Path(os.environ["SCRATCH"])

run_dir = scratch / "opensees" / "run01"

This allows your code to adapt automatically when run:

  • In JupyterHub

  • On Stampede3

  • Inside a Tapis job

Validating Paths Early#

Fail fast when files are missing:

if not input_file.exists():
    raise FileNotFoundError(f"Missing input: {input_file}")

This is especially important for long HPC jobs, where silent failures waste queue time.

Summary: Python Path Rules of Thumb#

  • Use pathlib.Path, not raw strings

  • Resolve paths before launching analyses

  • Prefer absolute paths for batch jobs

  • Use environment variables for portability

The notebooks in the Training Sections will build on these principles directly.

Contents