Module 4: Data Manipulation
Goal
Manipulate and prepare real-world data for analysis and machine learning.
This page focuses on the practical skills needed to turn raw, messy data into clean, structured inputs that models and analyses can actually use.
Estimated Time Impact
20-28 hours total
1. NumPy Foundations
NumPy: The Absolute Basics for Beginners
CodeSignal: NumPy Basics
Why it matters:
NumPy is the foundation under most Python data and machine learning work. Pandas, scikit-learn, PyTorch, and many scientific Python tools rely on the same ideas: arrays, shapes, indexing, vectorized operations, and numerical aggregation.
Before learners clean real datasets with Pandas, they should understand how numerical data is represented and transformed efficiently.
What to expect:
Learners will:
- Create and inspect NumPy arrays
- Understand array shape, dimension, size, and dtype
- Use indexing, slicing, boolean masks, and reshaping
- Apply vectorized operations instead of slow Python loops
- Compute basic statistics and aggregations
By the end of this section, learners will be ready to understand why Pandas operations feel array-like and how numerical data flows into machine learning workflows.
2. Pandas & Data Cleaning
Kaggle: Pandas Micro-Course
Why it matters:
Most real-world datasets are incomplete, inconsistent, and poorly formatted. Before any modeling or analysis can happen, data must be inspected, cleaned, and transformed.
Pandas is the primary tool used in Python for working with tabular data at scale, and it forms the backbone of most data pipelines.
What to expect:
Learners will:
- Load and explore large CSV and tabular datasets
- Filter, sort, and aggregate data efficiently
- Handle missing values and inconsistent formats
- Create new features from existing columns
- Apply vectorized operations for performance
By the end of this section, learners will be able to clean and manipulate real-world datasets with concise, readable code.
3. Basic Data Visualization with Matplotlib
CodeSignal: Introduction to Basic Plots with Matplotlib
Why it matters:
After learners clean and transform data, they need to inspect it visually. Basic plots help reveal trends, outliers, distributions, and relationships that are hard to catch from tables alone.
Matplotlib is also the lowest-level plotting foundation behind many Python visualization tools, so learning it early makes later Seaborn, Pandas plotting, and ML evaluation charts easier to understand.
What to expect:
Learners will:
- Create basic plots with Matplotlib
- Choose simple chart types for different data questions
- Label axes and titles clearly
- Use plots to explain patterns in cleaned data
- Build confidence before moving into advanced visualization libraries
By the end of this section, learners will be able to create simple visual checks for datasets before moving into SQL-heavy work or later machine learning evaluation.
4. SQL
SQL Tutorial - Essential SQL For The Beginners
Why it matters:
Most production data lives in relational databases. SQL is the universal language for querying, filtering, aggregating, and transforming data at scale.
Whether extracting features for machine learning or generating reports, SQL fluency is essential for working with real-world data systems.
What to expect:
Learners will:
- Write SELECT queries to retrieve and filter data
- Use JOIN operations to combine data from multiple tables
- Perform aggregations with GROUP BY and HAVING
- Create, update, and delete records with INSERT, UPDATE, DELETE
- Work with subqueries and common table expressions (CTEs)
- Understand indexing and query optimization basics
By completing this section, learners will be able to extract and manipulate data directly from databases used in production environments.
5. Software Engineering Tools
Git Branching Game
Why it matters:
Data work is collaborative and iterative. Without version control, experiments are lost, results are overwritten, and collaboration breaks down.
Git is the industry-standard tool for tracking changes, experimenting safely, and working in teams.
What to expect:
Learners will:
- Understand repositories, commits, and branches
- Save and version their work using git
- Push and pull changes from remote repositories
- Merge branches and resolve simple conflicts
- Develop safe workflows for experimentation
This section ensures learners can manage their code and data work reliably, both individually and in team settings.
Completion Checklist
- Completed NumPy: The Absolute Basics for Beginners and practiced array creation, indexing, slicing, reshaping, boolean masks, and basic aggregations.
- Completed at least one NumPy practice path: CodeSignal NumPy Basics, Keith Galli's NumPy Tutorial, or Patrick Loeber's NumPy Crash Course.
- Completed the Kaggle: Pandas Micro-Course, including all core lessons and exercises, and earned the official course certificate from Kaggle.
- Completed CodeSignal: Introduction to Basic Plots with Matplotlib and created basic plots from cleaned data.
- Completed the SQL Tutorial - Essential SQL For The Beginners (SQL Tutorial) and practiced all query types including JOINs, aggregations, and subqueries.
- Completed the W3Docs Git Basics Quiz (Git Basics Quiz) with a score of at least 85% and earned the Git certificate.
