Using the Ubuntu bash terminal on Windows 10 (installation instructions here), I installed Anaconda and have been using Jupyter notebooks without a problem. Unfortunately, Jupyter is unable to locate a runnable browser from within the subsystem, so I have to copy and paste the link it outputs in the terminal - but that is workable. The main issue comes when I try to open multiple notebooks. Normally, Jupyter would detect that a port (8888 by default) is already being used and make a new one, but it seems to fail to detect this so that when I use the link it generates, I end up looking at the first notebook I opened instead of the new one.
Any idea what the issue might be? And, if not, how I can manually get around this?
Assign different port number manually when you start the notebook. For example:
jupyter notebook --port=8889
try:
jupyter notebook --no-browser
I had similar issue with browser, I got
No web browser found: could not locate runnable browser.
I installed WSLU https://github.com/wslutilities/wslu.
Then I got
Start : This command cannot be run due to the error: The system cannot find the file specified.
At line:1 char:1
+ Start --h
+ ~~~~~~~~~
+ CategoryInfo : InvalidOperation: (:) [Start-Process], InvalidOperationException
+ FullyQualifiedErrorId : InvalidOperationException,Microsoft.PowerShell.Commands.StartProcessCommand
jupyter-notebook does not supply url as a parameter to wlsview. It passes a path with file to browser. eg
file:///home/myhome/.local/share/jupyter/runtime/nbserver-5058-open.html
with actual url
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<meta http-equiv="refresh" content="1;url=http://localhost:8888/tree?token=19b5f1fefb13f5fc315b05991175d1f8cb5ada9baaca6804" />
<title>Opening Jupyter Notebook</title>
</head>
<body>
<p>
This page should redirect you to Jupyter Notebook. If it doesn't,
<a href="http://localhost:8888/tree?token=19b5f1fefb13f5fc315b05991175d1f8cb5ada9baaca6804">click here to go to Jupyter</a>.
</p>
</body>
</html>
Create a file jupyter-notebook-browser with a content to extract actual url
#!/bin/bash
file=$(echo "$1" | sed 's/file:\/\///')
url=$(grep -oP 'href="\K([^"]*localhost[^"]+)' "$file")
wslview "$url"
then run jupyter-notebook --browser=jupyter-notebook-browser
or define BROWSER variable and run
export BROWSER="jupyter-notebook-browser"
jupyter-notebook
