Skip to content

Installation

Quick Install

pip install panelbox

That's it. PanelBox and all core dependencies will be installed automatically.

Requirements

Python: >= 3.9 (3.9, 3.10, 3.11, 3.12 supported)

Core dependencies (installed automatically):

Package Minimum Version Purpose
NumPy >= 1.24.0 Array operations
Pandas >= 2.0.0 Data handling
SciPy >= 1.10.0 Statistical functions
statsmodels >= 0.14.0 Econometric foundations
patsy >= 0.5.3 Formula parsing

Optional dependencies:

Package Minimum Version Purpose
matplotlib >= 3.5.0 Static plotting
plotly >= 5.0.0 Interactive charts
numba >= 0.56.0 Performance optimization

Installation Options

pip install panelbox

Upgrade to the latest version:

pip install --upgrade panelbox

git clone https://github.com/PanelBox-Econometrics-Model/panelbox.git
cd panelbox
pip install -e .

git clone https://github.com/PanelBox-Econometrics-Model/panelbox.git
cd panelbox
pip install -e ".[dev]"

Optional Extras

# Plotting support
pip install panelbox[plots]

# Performance optimization
pip install panelbox[performance]

# Development tools (testing, linting, formatting)
pip install panelbox[dev]

# Documentation tools
pip install panelbox[docs]

# Everything
pip install panelbox[all]

Verification

Verify that PanelBox is installed correctly:

import panelbox as pb

# Check version
print(f"PanelBox version: {pb.__version__}")

# Test with built-in dataset
data = pb.load_grunfeld()
print(f"Loaded {len(data)} observations from Grunfeld dataset")

# Quick model test
from panelbox import FixedEffects
model = FixedEffects("invest ~ value + capital", data, "firm", "year")
results = model.fit()
print(f"R-squared (within): {results.rsquared_within:.4f}")

Expected output:

PanelBox version: 1.x.x
Loaded 200 observations from Grunfeld dataset
R-squared (within): 0.7668

Jupyter and Colab Setup

Local Jupyter

# Install in your Jupyter environment
pip install panelbox

# If using a virtual environment, register the kernel
python -m ipykernel install --user --name=panelbox_env

Google Colab

In the first cell of your notebook:

!pip install panelbox -q
import panelbox as pb
print(pb.__version__)

Colab Notebooks

PanelBox tutorials are available as ready-to-run Colab notebooks. See the User Guide sections for links.

Virtual Environments

We recommend using a virtual environment to avoid dependency conflicts:

python -m venv panelbox_env
source panelbox_env/bin/activate   # Linux/macOS
panelbox_env\Scripts\activate      # Windows
pip install panelbox

conda create -n panelbox_env python=3.11
conda activate panelbox_env
pip install panelbox

Troubleshooting

ModuleNotFoundError: No module named 'panelbox'

Ensure PanelBox is installed in the active Python environment:

pip list | grep panelbox

If not found, install it. If using Jupyter, make sure the notebook kernel matches the environment where PanelBox is installed.

Dependency Conflicts

Create a fresh virtual environment:

python -m venv fresh_env
source fresh_env/bin/activate
pip install panelbox

Windows: DLL Load Failed

Install the Microsoft Visual C++ Redistributable from Microsoft's download page.

macOS Apple Silicon (M1/M2/M3)

Use conda with native ARM64 support:

conda create -n panelbox_env python=3.11
conda activate panelbox_env
pip install panelbox

Performance Warning: Numba Not Available

PanelBox works without Numba, but large datasets benefit from it:

pip install numba

System Information for Bug Reports

import sys, platform
import panelbox as pb

print(f"Python:    {sys.version}")
print(f"Platform:  {platform.platform()}")
print(f"PanelBox:  {pb.__version__}")

Include this output when reporting issues.

Next Steps