I connect to my own jupyter server via vpn/ssh with vscode. Jupyter server runs as a daemon.
Having asked gpt, read articles here and tried many config versions, I still end up doing this at the beginning of each notebook:
BASE_PATH = '/home/bartsimpson/dev/'
import sys
sys.path.append(BASE_PATH)
Is there a way to set it somewhere in my jupyter server or in vscode so that I could easily do this without setting path manually?
from library.myclass import MyClass
if it resides in /home/bartsimpson/dev/library/myclass.py
?
P.S. Setting neither PYTHONPATH
, nor JUPYTER_PATH
prior to launch helps.
You could go to the directory of the kernel you're using which involves the kernel.json
file.
In the same directory, create a new file called ipython_kernel_config.py
.
c = get_config()
c.InteractiveShellApp.exec_lines = [
"import sys",
"sys.path.append('/home/bartsimpson/dev/')",
]
Now, each time you start a Jupyter notebook using this kernel, the lines in exec_lines
will be executed