This repo contains different Colab Notebooks used to generate isoscape models.
The steps are detailed at go/github, what follows is a summary:
git
to use your corp email
address: git config --global user.email "email@example.com"
Please consult this document for known issues. Feel free to report one by adding a new row.
http://go/ddf-github-known-issues
NOTE: Googlers might be redirected to the internal Colab, so check that the URL matches the one above. Otherwise, there should be a toggle near the top that reads
Switch to prod
.
From the window choose GitHub tab and select Include private repos
.
Enter your GitHub username.
main.ipynb
should show up; if it does not, refresh the tab or open a new
one, making sure that you are connected to your account. Open it, and follow
next steps.
Please use the official docker container for running a local runtime.
See the official colab instructions
for more details.
Once the server has started, it will print a message with the initial backend URL used for authentication. Make a copy of this URL as you'll need to provide this in the next step.
In Colab, click the Connect
button and
select Connect to local runtime...
. Enter the URL from the previous step in
the dialog that appears and click the Connect
button. After this, you should
now be connected to your local runtime.
GDrive is currently not supported in local runtimes, but can be used in hosted runtimes.
After connecting to colab,
click Connect > Connect to a hosted runtime
(you may also just
click Connect
).
In the Parent Directories cell, set USE_GDRIVE = True
. If using your
own GDrive, set USE_SHARED_GDRIVE = False
and
set PERSONAL_GDRIVE_PARENT_DIR
to the name of the folder containing all the
data. Otherwise, set USE_SHARED_GDRIVE = True
and
set SHARED_GDRIVE_PARENT_DIR
to the folder in the shared drive containing
all the data, prefixed by the shared drive name.
Run as normal. When you reach the Mount GDrive cell, grant Colab permission to access GDrive.
One option is to use VSCode, which has Jupyter support. Googlers see go/vscode/install.
Intellij also has Jupyter support.
Remember to exclude working directories from backups.
Consider enabling Rich Jupyter Notebook Diffs via Feature Previews to make Notebook diffs easy to read.