Body
PyCharm is installed on all our Ubuntu 16 workstations.
- Change your shell to bash.
- Logout, then log back in to ensure you pick up the changes.
- Download Anaconda. Save it in
/local-scratch for faster access.
- Run the installer by typing
bash /local-scratch/Anaconda_installer_whateverversion.sh.
- When prompted for an installation directory, enter
/local-scratch/myanaconda (or any other directory name that does not yet exist).
- Conda should prompt you to activate it. Answer “yes”. This should not require a password. If you
[sudo] asks for your password, you are probably out of disk space or you have exceeded your home directory quota. See our Disk Quota page for troubleshooting tips.
- Use
conda install to obtain OpenCV, cudatoolkit, tensorflow, scikit-learn or whatever else you need.
- Point PyCharm at Anaconda’s
bin/python2.7 or bin/python3.6 (see below).
Warning
/local-scratch is not backed up. Keep backups of important work in your shared lab file space or SFU Vault.
TELLING PYCHARM TO USE ANACONDA PYTHON (NEW PROJECTS)
When creating a new project, the “Select Python Interpreter” dialog box will appear. Fill in the path to your Anaconda Python binary, e.g. /local-scratch/anaconda/bin/python2.7 and click [OK].
TELLING PYCHARM TO USE ANACONDA PYTHON (EXISTING PROJECTS)
With an existing project open, choose the File menu => Settings…* => Project: YourProject => [gear icon] => **Add Local.
Fill in the path to your Anaconda Python binary, e.g. /local-scratch/anaconda/bin/python2.7 and click [OK]. The first time you do this, PyCharm will spend several minutes scanning available modules.
