jupyter-alloy-notebooks-retro

⚠️ This is a purpose-built backport. You almost certainly want the main package instead.

It exists to run Alloy Notebooks inside a legacy classic Jupyter Notebook 6.x environment (a notebook 6.5.x + RISE 5.7 presentation setup). Classic Notebook 6 cannot load JupyterLab extensions, so this is a from-scratch nbextension port, maintained on a best-effort basis.

If you are using JupyterLab 4 (or modern Jupyter in general), do not install this. Use the real extension instead:

pip install alloy-notebooks

(source: Columbia-Data-Labs/alloy-notebooks)

What it does

• SQL, DuckDB and R cells — pick a language from the unified cell dropdown. The cell's source is never rewritten; only what is sent to the kernel is.
• Result tables and charts — Table/Chart toggle, an opt-in chart per column computed over the full result, click-to-rename chart titles, and Configure Chart (bar, line, scatter, pie, histogram, area).
• Connection manager — a slide-over panel (toolbar database button or the Alloy menu) to save, edit and connect to SQL Server, PostgreSQL, MySQL, SQLite and DuckDB. Live connections are restored after a kernel restart.
• Query indicator, per-cell highlighting and language icons.
• Everything else from the kernel side — alloy_connect, %%alloy_sql, %alloy_to_r and the other magics come from the alloy-notebooks dependency.

Notebooks round-trip with the JupyterLab extension: both use the same cell metadata (alloy:language, alloy:resultView, alloy:chartConfig, alloy:chartShown, alloy:chartTitles), the same result MIME type, and the same saved connections in ~/.alloy/connections.json.

During a RISE slideshow the connection panel and language icons are hidden.

Install

conda activate rise6
pip install jupyter-alloy-notebooks-retro

Then restart jupyter notebook. The nbextension and server extension enable themselves at install time. Installed alongside it, alloy-notebooks also brings jupysql, plotly and their dependencies.

DuckDB and R cells need their engines, as they do in the main package:

pip install duckdb      # DuckDB cells
pip install rpy2        # R cells (R itself must also be installed)

To uninstall:

pip uninstall jupyter-alloy-notebooks-retro

Requirements

• notebook >= 6, < 7 (Notebook 7 is Lab-based and is served by the main package)
• Python 3.10+
• A current browser (Chrome/Edge)

Differences from the JupyterLab extension

• The language-icon preference is stored in the browser's localStorage, because classic Notebook has no Settings Registry.
• The core cell-type dropdown is hidden in favour of the unified one, as in Lab. With several cells selected, the dropdown applies to all of them.

License

BSD-3-Clause, same as the parent project.