Principal Component Analysis tutorial

Overview

What is Principal Component Analysis?

Principal Component Analysis is a unsupervised learning technique. Rotates data onto directions that preserve the most variance. Its central idea is summarized by Z = X̃Wₖ.

This guide connects theory to practice. You will trace the input, intermediate process, and output; decode the notation; run a dependency-light implementation; then repeat the workflow with scikit-learn and evaluate the result.

Preparation

Prerequisites and tools

You do not need an advanced software stack. Start with a recent Python environment and the fundamentals below, then install only the packages used by the practical example.

• Comfort with Python functions, NumPy arrays, and basic descriptive statistics.
• A clear distinction between training data, validation data, and untouched test data.
• Familiarity with features, targets, preprocessing, and task-appropriate evaluation metrics.

Process

How Principal Component Analysis works

1. 1

   Input

   Correlated numeric features.

2. 2

   Prepare and configure

   Check shapes, value ranges, ordering assumptions, missing values, and the parameters that control the algorithm’s behavior.

3. 3

   Algorithm process

   Find covariance eigenvectors ordered by explained variance.

4. 4

   Output

   Lower-dimensional principal-component coordinates.

5. 5

   Validate

   Choose component count from cumulative explained variance and downstream validation performance.

Core concept

Formula and intuition

The notation captures the main operation or complexity statement behind Principal Component Analysis. Read it together with the symbol key above and the step-by-step process in this guide.

Applications

When to use Principal Component Analysis

Visualization, compression, and noise reduction.

Implementation

Principal Component Analysis from scratch in Python

This dependency-light example emphasizes the algorithm’s mechanics so each important step remains visible.

import numpy as np

rng = np.random.default_rng(0)
X = rng.normal(0, 1, (200, 5))
X_centered = X - X.mean(axis=0)  # PCA requires zero-mean data

# eigenvectors of the covariance matrix point along directions of maximum variance
cov = np.cov(X_centered, rowvar=False)
eigvals, eigvecs = np.linalg.eigh(cov)
order = np.argsort(eigvals)[::-1]   # sort directions by variance explained, descending
components = eigvecs[:, order[:2]]  # keep the top 2 directions

projected = X_centered @ components  # coordinates of each point in the new, smaller space
explained_variance = eigvals[order[:2]] / eigvals.sum()

Practical tutorial

Build Principal Component Analysis with scikit-learn

PCA computes stable components and exposes explained-variance ratios for selecting a useful reduced dimension.

1. Prepare the data.Correlated numeric features. Validate its shape, type, range, and ordering before training or execution.
2. Configure the algorithm.Begin with explicit, conservative parameters and a fixed random seed whenever the library supports one.
3. Fit or execute.Find covariance eigenvectors ordered by explained variance.
4. Inspect the result.Lower-dimensional principal-component coordinates. Then apply the evaluation checks in the next section.

from sklearn.datasets import load_digits
from sklearn.decomposition import PCA
from sklearn.preprocessing import StandardScaler

X, _ = load_digits(return_X_y=True)
X_scaled = StandardScaler().fit_transform(X)
pca = PCA(n_components=0.95, svd_solver="full")
Z = pca.fit_transform(X_scaled)
print("original dimensions:", X.shape[1])
print("retained dimensions:", Z.shape[1])
print("variance retained:", pca.explained_variance_ratio_.sum())
X_reconstructed = pca.inverse_transform(Z)
print("reconstruction MSE:", ((X_scaled - X_reconstructed) ** 2).mean())

API details and version-specific options: official scikit-learn reference →

Evaluation

How to evaluate the result

A successful run is not enough. Evaluate the output against the intended use, compare it with a simple baseline, and preserve a genuinely unseen test case whenever the task involves learned parameters.

• Choose component count from cumulative explained variance and downstream validation performance.
• Check reconstruction error when dimensionality reduction is used for compression.

Common mistakes

Pitfalls and how to avoid them

These failure modes are common in tutorials and production systems. Treat them as review questions, not just after-the-fact debugging advice.

• PCA captures variance, not necessarily task-relevant information.
• Results are sensitive to feature scale unless units are already comparable.
• Component signs are arbitrary and do not change the represented subspace.

Project checklist

Before using Principal Component Analysis in a project

• ✓Define the prediction or discovery objective before selecting the algorithm.
• ✓Split data before fitting preprocessing and tune only inside cross-validation.
• ✓Record data versions, random seeds, features, hyperparameters, and evaluation metrics.
• ✓Compare against a simple baseline and inspect errors by meaningful subgroups.
• ✓Monitor input drift and real-world performance after deployment.

Further reading

Official documentation and next steps

This guide is an educational introduction, not a substitute for domain validation. For consequential applications, review the source documentation, test against representative data, and involve a subject-matter expert.

Continue learning

Related machine learning guides

• Linear RegressionFits a straight line that minimizes squared prediction errors.
• Logistic RegressionMaps a weighted feature score to a class probability.
• Decision TreeSplits the feature space into rule-based regions.