Installation

Install from PyPI (pip)

The simplest way to install bnlearn is using pip:

pip install bnlearn

To force install the latest version, use the -U (update) argument:

pip install -U bnlearn

Install from GitHub

To install the latest development version directly from GitHub:

pip install git+https://github.com/erdogant/bnlearn

Create Environment

For better dependency management, it’s recommended to install bnlearn in an isolated Python environment using conda:

conda create -n env_bnlearn python=3.10
conda activate env_bnlearn

Create a new conda environment.

After activation, your command prompt should show the environment name. For example:

(env_bnlearn) D:\>

Uninstall

To remove the bnlearn installation and its environment:

# List all active environments
conda env list

# Remove the bnlearn environment
conda env remove --name env_bnlearn

# Verify removal by listing environments again
conda env list

Validate Installation

To verify your installation, start Python in your console:

python

Then run the following code, which should generate a figure:

import bnlearn as bn
df = bn.import_example()
model = bn.structure_learning.fit(df)
G = bn.plot(model)

Troubleshooting Import Errors

If you’re using Jupyter Notebook or Google Colab, you might encounter a NumPy version compatibility error:

import bnlearn as bn
# Error message:
RuntimeError: module compiled against API version 0x10 but this version of numpy is 0xf
ImportError: numpy.core.multiarray failed to import

This error occurs because bnlearn requires NumPy version 1.24.1 or higher. To resolve this:

1. To fix this, you need an installation of numpy version=>1.24.1 which is installed during the bnlearn installation. However, when you are using colab or a jupyter notebook, you need to reset your kernel first to let it work.

2. If using Colab or Jupyter Notebook: - Go to the menu - Click Runtime -> Restart runtime - Re-import bnlearn

---

---