Installing Python and Data Science Environment

# CHAPTER 2

Installing Python and Data Science Environment #

1. Chapter Introduction #

2. The Anaconda Distribution (Recommended) #

Python on its own is just a bare-bones programming language. If you install standard Python from python.org, you will have to manually install Pandas, NumPy, and Jupyter one by one via the command line.

Anaconda is the industry-standard distribution for Data Science. It is a massive package that pre-installs Python *and* the top 100 data science libraries simultaneously.

Step-by-step Installation:

1. 1. Go to anaconda.com/download.

1. 2. Download the installer for Windows, Mac, or Linux.

1. 3. Run the installer. Leave all settings on their defaults.

1. 4. Once installed, search your computer for Anaconda Navigator and open it.

3. Launching Jupyter Notebook #

Jupyter Notebook is the primary tool you will use to write Python code in this course. It runs in your web browser but processes data locally on your computer.

1. 1. Open Anaconda Navigator.

1. 2. Find the tile labeled Jupyter Notebook.

1. 3. Click Launch.

1. 4. A black terminal window will open (do not close this!). Then, a new tab will open in your web browser (usually at localhost:8888). This is your Jupyter file browser.

4. VS Code: The Professional IDE #

While Jupyter Notebook in the browser is great for beginners, most professional Data Scientists and Machine Learning Engineers eventually migrate to Visual Studio Code (VS Code). It is a powerful, free code editor built by Microsoft.

How to integrate VS Code with Data Science:

1. 1. Download VS Code from code.visualstudio.com.

1. 2. Open VS Code, go to the Extensions tab (square boxes icon on the left).

1. 3. Search for and install the Python extension.

1. 4. Search for and install the Jupyter extension.

1. 5. You can now create files ending in .ipynb and run Jupyter notebooks directly inside VS Code!

5. Managing Libraries with pip and conda #

If you need a library that Anaconda didn't install, you must download it from the internet using a package manager. You have two options: pip (Python's default) and conda (Anaconda's default).

To install a package, open your Anaconda Prompt (Windows) or Terminal (Mac) and type:

# Using pip
pip install seaborn

# Using conda
conda install seaborn

*Note: In Jupyter Notebook, you can run these commands directly in a code cell by adding an exclamation mark: !pip install seaborn.*

6. Virtual Environments #

Imagine Project A needs Pandas version 1.0, but Project B requires Pandas version 2.0. If you only have one installation of Python on your computer, you can't run both projects.

Virtual Environments solve this. They create an isolated, "fenced-off" version of Python for each project.

Creating an environment with Conda:

# Create a new environment named 'my_ds_project' with Python 3.10
conda create --name my_ds_project python=3.10

# Activate (enter) the environment
conda activate my_ds_project

# Now any pip installs only affect THIS environment!
pip install pandas

7. Common Mistakes #

• Installing Multiple Pythons: Beginners often install Python from python.org, then install Anaconda, then install Python from the Windows Store. This creates massive conflicts. Choose ONE ecosystem (Anaconda is best) and stick to it.

• Closing the Jupyter Terminal: When you launch Jupyter, a scary-looking black terminal window stays open in the background. If you close it, your notebook will disconnect and stop working. Minimize it, don't close it!

8. MCQs #

9. Interview Questions #

• Q: Explain the concept of Virtual Environments. Why are they a critical best practice in professional data science teams?

• Q: What is the difference between writing code in a standard .py script versus a .ipynb Jupyter Notebook?

10. Summary #

11. Next Chapter Recommendation #