Troubleshooting#

Common issues#

Are you sure you’re using the fetch3 conda environment? (Remember that if you open a new terminal, you need to activate the environment again)
Depending on the terminal you’re using, you should probably see fetch3-dev in parentheses somewhere in the command line. If you see base in parentheses, this means you’re using the base conda environment instead of the fetch3 environment.
- You can run conda env list to check which environment is active (there is a * next to the active environment)
- If you’re running one of the .ipynb notebooks to look at the model output, make sure that the notebook is using the fetch3-dev environment
If you’re sure you’re using the fetch3-dev environment and still getting a ModuleNotFoundError, see Installation errors

Are you sure you’re using the correct directory?
- For example, if you get this error when you run python main.py, run ls and check that there is a main.py file in the current directory.
- If you change the directories in the notebooks, double-check that you’re giving it the correct directory.
- If you changed the output path for the model, make sure you specified an output directory that exists on your computer.
- Check whether a full filepath or a directory was supposed to be provided, and that you provided the correct one (e.g., if the error was with the path to the config file, did you provide the full filepath including the filename and extension, or just a directory?)
For directory errors in Windows:
- If you’re sure you’re using the correct directory and still getting an error, it’s probably an issue with parsing the Windows path correctly. There might be characters (like spaces) that need to be escaped for the path to be parsed properly.
  - Try using double backslashes in the file paths (i.e., change \ to \\)
  - Try keeping things in directories that don’t have any special characters or spaces in the path

For Windows, make sure git is installed before you install the fetch3 conda environment
If you are using the fetch3 environment and are still getting a ModuleNotFoundError, check that the module in question was actually installed in the fetch3-dev environment. With the fetch3-dev environment active, run conda list to see the packages that are installed in fetch3-dev.
If you find that the fetch3 environment is missing packages that should have been installed, try removing and re-installing the environment.
- To remove the fetch3 environment:
```
conda deactivate
conda remove -n fetch3-dev --all
```
- Next, make sure you have the latest version of the FETCH3 code, and re-install the fetch3-dev environment. See Install FETCH3’s dependencies

If you’re getting an error about not having permission to access files in Windows, you might need to change your permissions settings.

Todo

Add instructions about changing permissions settings in powershell to run as admin by default

If you’re having an issue not covered in the above section, make sure you’ve updated everything to the latest version.

Make sure you have the latest version of the FETCH3 code from GitHub
Using the most recent version of the FETCH3 repository, update your conda environment. With the fetch3-dev environment active, run:
```
conda env update -file fetch3_requirements.yml --prune
```

Update boa:

pip install -U git+https://github.com/madeline-scyphers/boa.git

If you’re still having issues, please submit a GitHub issue