Install Phoenix Arize in Python

Phoenix Arize is a powerful open-source tool for observability. It helps you monitor and debug your machine learning models. Installing it correctly in Python is the first step. This guide will show you how to do it quickly.

We will cover the installation process step-by-step. You will see clear code examples. We will also discuss common issues. By the end, you will have Phoenix Arize running in your Python environment.

What Is Phoenix Arize?

Phoenix Arize is a library for LLM observability. It tracks prompts, responses, and model performance. It helps you find errors and improve your AI applications. You can use it with frameworks like LangChain or LlamaIndex.

This tool is essential for production AI systems. It gives you insights into model behavior. You can detect drift, hallucinations, and latency issues. Installing it is the foundation for all this functionality.

System Requirements

Before you install, check your system. Phoenix Arize works with Python 3.8 or higher. You need pip installed. A virtual environment is recommended for clean setup.

Your system should have at least 1GB of RAM. A stable internet connection is needed for downloading packages. No special hardware is required.

Step 1: Create a Virtual Environment

Always use a virtual environment. This keeps your project dependencies isolated. It prevents conflicts with other Python projects.

Open your terminal or command prompt. Navigate to your project folder. Run the following command:

# Create virtual environment
python -m venv phoenix-env

This creates a folder named phoenix-env. Now activate it. On Windows:

phoenix-env\Scripts\activate

On macOS or Linux:

source phoenix-env/bin/activate

Your terminal will show the environment name. This means you are inside the virtual environment.

Step 2: Install Phoenix Arize

Now install the main package. Use pip to install arize-phoenix. This is the core library.

# Install Phoenix Arize
pip install arize-phoenix

This command downloads and installs all dependencies. The process may take a few minutes. You will see progress bars and status messages.

For LLM observability, you may need extra packages. Install the LLM support:

# Install with LLM support
pip install arize-phoenix[llama-index]

Replace llama-index with your framework. Options include langchain or openai. This adds specific integrations.

Step 3: Verify the Installation

Check if Phoenix Arize installed correctly. Open a Python shell or create a test script. Run this code:

# Verify installation
import phoenix as px
print(px.__version__)

You should see a version number. For example:

3.0.0

If you see an error, check the next section. Common issues include missing dependencies or Python version problems.

Step 4: Start the Phoenix Server

Phoenix Arize runs a local server for the UI. Start it with this command:

# Start Phoenix server
import phoenix as px
px.launch_app()

You will see output like:

🌍 Running on http://127.0.0.1:6006

Open this URL in your browser. You will see the Phoenix dashboard. It is empty now. You will populate it with data later.

Step 5: Use a Simple Example

Let's test Phoenix with a basic trace. This example logs a prompt and response. Create a new Python file:

# simple_trace.py
import phoenix as px
from phoenix.trace import Span

# Start the app
px.launch_app()

# Create a simple span
span = Span(
    name="test-span",
    attributes={"prompt": "Hello, world!", "response": "Hi there!"}
)

# Log the span
px.log(span)

print("Trace logged successfully!")

Run the file:

python simple_trace.py

Check the Phoenix dashboard at http://127.0.0.1:6006. You should see your trace. This confirms everything works.

Common Installation Issues

Sometimes installation fails. Here are solutions for common problems.

Problem: pip not found. Install pip first. On most systems, use python -m ensurepip. Or upgrade pip with python -m pip install --upgrade pip.

Problem: Permission errors. On Linux or macOS, use sudo before pip. Or install in a virtual environment without sudo. The virtual environment method is safer.

Problem: Dependency conflicts. Use a fresh virtual environment. List all packages with pip list. Remove conflicting packages if needed.

Problem: Python version too old. Update Python to 3.8 or newer. Download from python.org. Check your version with python --version.

Installing for Different Frameworks

Phoenix Arize supports multiple frameworks. Install specific integrations as needed.

For LangChain:

pip install arize-phoenix[langchain]

For OpenAI:

pip install arize-phoenix[openai]

For LlamaIndex:

pip install arize-phoenix[llama-index]

Each framework has its own integration. Check the Phoenix documentation for details. These extras add tracing capabilities.

Upgrading Phoenix Arize

Keep Phoenix updated for new features. Use pip to upgrade:

pip install --upgrade arize-phoenix

This updates the core package. Upgrade extras separately:

pip install --upgrade arize-phoenix[langchain]

Check the version after upgrade. Run px.__version__ again. Ensure you have the latest stable release.

Uninstalling Phoenix Arize

If you need to remove Phoenix, use pip:

pip uninstall arize-phoenix

This removes the main package. Uninstall extras separately with similar commands. Delete the virtual environment folder to clean everything.

Conclusion

Installing Phoenix Arize in Python is straightforward. Use a virtual environment for safety. Install the core package with pip. Verify with a simple import test. Start the server and log your first trace.

This tool gives you powerful observability for AI models. It helps you debug and monitor performance. With this installation guide, you are ready to explore Phoenix Arize fully.

Remember to check the official documentation for advanced features. Experiment with different frameworks. Your AI applications will benefit from this essential tool.