Uploading external notebooks

You can upload Jupyter Notebooks (.ipynb or .zip extension) that were created in a Jupyter environment outside of IBM Cognos Analytics.

The uploaded notebooks are matched against the following JSON schemas:

https://github.com/jupyter/nbformat/blob/master/nbformat/v3/nbformat.v3.schema.json (version 3 (v3) notebooks)
https://github.com/jupyter/nbformat/blob/master/nbformat/v4/nbformat.v4.schema.json (version 4 (v4) notebooks)

If a notebook doesn't conform to either of these schemas, the upload is rejected.

Jupyter Server creates all new notebooks using the v4 schema.

About this task

The process of uploading notebook files is similar to the process of uploading spreadsheets and text files that are used as data sources in Cognos® Analytics. For example, the notebook files have the same size limitations as other uploaded file types. For more information, see Upload data files.

Procedure

Use the following methods to upload the notebooks:

From the Open menu , click Upload data. Locate the notebook files (.ipynb or .zip) on your local drive or on the LAN, and select one or multiple files to upload them.
In the welcome page, drag one or multiple .ipynb or .zip files from your local drive onto the welcome page to activate the Quick upload functionality. When Quick upload appears, drop the files into the Notebook box.
From Team content or My content, click Upload data. Locate the notebook files on your local drive or on the LAN, and select one or multiple files to upload them. The files are saved to the folder from which you initiated the upload.

Results

A notebook is created for each uploaded .ipynb file in My content. However, when the upload was initiated from a specific folder, the notebooks are created in that folder.

What to do next

You can seamlessly open an uploaded v4 notebook in the edit mode. To open a v3 notebook in the edit mode, the Jupyter server temporarily converts it to the v4 format. If you save the notebook, it is saved in the v4 format.

Tip: Older versions of Jupyter might not read the v4 format. To preserve the original notebook version, close it without saving.

You can view both v3 and v4 notebooks in the view-only mode. Because you can't interact with notebooks in this mode, there is no need to convert v3 notebooks to v4.

You can import uploaded v3 and v4 notebooks into a Cognos Analytics dashboard. The Notebook widget in the dashboards conforms to the v4 JSON schema so the visualizations in v4 notebooks display seamlessly in the dashboard. To ensure that the v3 notebook visualizations display properly in the dashboard, you must open and save the v3 notebooks in the edit mode before you import them into the dashboard.