Setting Up Your Python Environment for Scientific Computing: A Step-by-Step Guide

Who Is This For?

You are a graduate student in biology, chemistry, or materials science. You have heard that Python can help with your data, but you have never set up a programming environment. This guide walks you through every step—no prior experience assumed.

By the end, you will have a working Python setup with Jupyter notebooks, essential scientific packages, and VS Code as your editor.

Step 1: Install Miniconda

Why Miniconda instead of Anaconda? Miniconda is lightweight (~50 MB vs ~500 MB), installs faster, and gives you full control over which packages you need.

1. Go to docs.conda.io/en/latest/miniconda.html
2. Download the installer for your OS (Windows: .exe, macOS: .pkg, Linux: .sh)
3. Run the installer. Important: check "Add Miniconda to PATH" on Windows
4. Verify installation:

conda --version
# Expected: conda 24.x.x

Step 2: Create Your First Environment

Environments isolate project dependencies. Think of them as separate toolboxes—one for AFM analysis, one for MD simulations, one for ML projects.

# Create an environment with Python 3.10 (most packages support this)
conda create -n matsci python=3.10 -y

# Activate it
conda activate matsci

# Your prompt should now show (matsci)

Step 3: Install Essential Packages

For materials/biomaterials research, here are the packages you need right away:

# Core scientific stack
conda install numpy scipy pandas matplotlib jupyter -y

# Additional useful libraries
conda install -c conda-forge seaborn scikit-learn scikit-image -y

# For atomic/molecular simulations (install as needed)
# conda install -c conda-forge ase pymatgen mdanalysis -y

Verify everything works:

# test_env.py
import numpy as np
import scipy
import pandas as pd
import matplotlib.pyplot as plt
import sklearn

print(f"NumPy: {np.__version__}")
print(f"SciPy: {scipy.__version__}")
print(f"Pandas: {pd.__version__}")
print(f"scikit-learn: {sklearn.__version__}")
print("All packages OK!")

Step 4: Set Up Jupyter Notebook

Jupyter is the interactive coding environment most scientists use for exploration:

# Start Jupyter in your project folder
cd path/to/your/project
jupyter notebook

# Or use JupyterLab (more modern interface)
jupyter lab

Pro tip: Create a shortcut command for quick launch. On Windows, create jupyter_lab.bat:

@echo off
call conda activate matsci
jupyter lab

Step 5: VS Code for Serious Development

When notebooks get too messy (and they will), switch to VS Code:

1. Download from code.visualstudio.com
2. Install the Python extension (Microsoft)
3. Open your project folder: File > Open Folder
4. Select your conda environment: Ctrl+Shift+P > "Python: Select Interpreter" > choose matsci

Essential VS Code extensions for scientists:

• Python (Microsoft): IntelliSense, debugging
• Jupyter (Microsoft): Notebook support inside VS Code
• Rainbow CSV: Color-coded CSV files
• GitLens: Git integration

Step 6: Package Management Reference

# List installed packages
conda list

# Install a specific version
conda install numpy=1.24.3

# Install from conda-forge (community-maintained channel)
conda install -c conda-forge package_name

# Use pip when conda does not have the package
pip install package_name

# Export your environment (for sharing/reproducibility)
conda env export > environment.yml

# Recreate from file
conda env create -f environment.yml

Step 7: Common Problems and Solutions

Problem	Solution
---------	----------
conda not recognized	Restart terminal or reinstall with "Add to PATH" checked
Package install hangs	Use conda install -c conda-forge instead of defaults
DLL load failed on Windows	Install Visual C++ Redistributable
Jupyter kernel not found	python -m ipykernel install --user --name matsci
Conda slow on Windows	Use mamba instead: conda install mamba -c conda-forge then mamba install ...

Step 8: Quick Start Template

Save this as start_project.bat (Windows) or start_project.sh (macOS/Linux):

# Windows: start_project.bat
conda activate matsci
jupyter lab

Double-click to start working. That is it—you are ready to code.

References

• Conda documentation: docs.conda.io
• Jupyter documentation: docs.jupyter.org
• VS Code Python setup: code.visualstudio.com/docs/python/python-tutorial
• Scientific Python ecosystem: scientific-python.org