Variable Explorer for JupyterLab

A Spyder/Stata-grade variable explorer for JupyterLab 4. Opens in its own window so you can put it on a second monitor.

Install

pip install variable-explorer

That's it. Restart JupyterLab if it's running.

Usage

1. Open a notebook in JupyterLab
2. Click the Variables button in the notebook toolbar (spreadsheet icon)
3. The Variable Explorer opens in a separate browser window
4. Run cells — the variable list updates automatically
5. Click any DataFrame, list, or dict to view its contents

Features

• Detached window — lives outside JupyterLab, move it to another monitor
• AG Grid — fast virtual scrolling, handles large DataFrames
• Column sorting — click headers to sort (kernel-side, fast even on millions of rows)
• Inline histograms — distribution charts in every column header with hover tooltips
• Conditional formatting — numeric heatmaps, boolean color-coding
• Metadata sidebar — column types, null counts, stats (like Stata's Variable Manager)
• Properties panel — dataset shape, memory usage, dtypes
• Cell editing — double-click to edit values, synced back to the kernel
• Resizable panels — drag the dividers between panels
• Auto-refresh — variable list updates after every cell execution
• Supports many types — DataFrames, Series, numpy arrays, lists of dicts, dicts of lists, nested structures

Requirements

• JupyterLab 4.x
• Python 3.10+
• pandas (for DataFrame viewing)

Development

# Clone and install in development mode
git clone https://github.com/variable-explorer/variable-explorer.git
cd variable-explorer
pip install -e .

# Rebuild after making changes
jlpm install
jlpm build
pip install -e .

Requires Node.js 18+ for development builds.