Troubleshooting#
Check the active environment#
Run these commands in the terminal where installation or execution failed:
python --version
python -m pip --version
python -m pip check
Use Python 3.12 or later. If pip reports conflicting dependencies, resolve them or retry in a fresh environment. Do not ignore the conflict and assume the installation is complete.
For a pip installation error, update pip in that environment and retry:
python -m pip install --upgrade pip
python -m pip install autolens
Missing files when running examples#
Run scripts from the workspace root, not from inside a scripts subfolder:
cd path/to/autolens_workspace
python scripts/imaging/start_here.py
Replace path/to/autolens_workspace with your workspace location. Check that the dataset required by the example is present; a missing file is not always a working-directory problem.
Workspace and package versions#
A checkout of the workspace’s main branch can be newer than your installed library. Check your version:
python -c "import autolens; print(autolens.__version__)"
Use a workspace release tag matching that version where available, or update to a compatible published library/workspace pair. Do not change API calls just to work around an unexplained version mismatch.
Plot windows do not open#
Check whether your Python session has access to a graphical desktop. On a remote or headless machine, save figures to files instead of opening windows.
The workspace setting is config/visualize/general.yaml → general → backend. See Matplotlib’s backend guide before choosing a backend; interactive backends require their corresponding GUI packages.
GPU not detected#
See GPU setup and check the output of jax.devices().
Older Python versions#
Use a new environment with Python 3.12 or later for current releases. If you need a historical version, pin that version explicitly and use its corresponding documentation.
Get help#
Post in PyAutoLabs Discussions with your operating system, Python and package versions, the command you ran, and the complete traceback.