Setup_Remote Interpreter

In PyCharm it is possible to run your program on a remote interpreter instead of your local system. This is useful if you have a server where your necessary environment is already set up. This is the case for us, because on the lsdf cluster, we have a working Ginkgo and SSGet installation.

Necessary steps

In PyCharm

• open PyCharm
• go to Tools → Deployment → Configuration
• type SFTP
• host: lsdf-28-131.scc.kit.edu
• port: 22
• your username: u****
• authentication: your password
• root path: autodetect
• Done

Next:

• go to File → Setting → Project → Project Interpreter
• click the wheel and selet Add
• select ssh interpreter
• select existing server config
• select the server you just created in the previous steps
• choose create copy
• choose interpreter /usr/bin/python3.6
• leave settings for sync folder
• check the box next to autmatically upload

Now you are done setting up PyCharm and it should start a process that uploads your project to the server.

On the server

The last thing to do is installing all dependencies on the server. You most likely won't have root access and therefore need a workaround to install packages.

• open a terminal and ssh into the server.
• look for the folder tmp.

In this folder should be a folder called python_project_ followed by same number. This is where your project is uploaded to.

• open this folder
• enter the command

wget https://bootstrap.pypa.io/get-pip.py && python3.6 get-pip.py --user

python3.6 -m pip install -r requirements.txt --user

This will allow you install all necessary dependencies without having root access on the server.