Computing and Networks
Data Science with Python
From raw data to models and insight
AI generated by Nikhil Kumar Rajput
14 chapters
Contents¶
Front matter
Part I · Foundations
Chapter 2: Python Essentials for Data Work
Chapter 3: Numerical Computing with NumPy
- 3.1 Why arrays?
- 3.2 Creating arrays
- 3.3 Indexing and slicing
- 3.4 Boolean masks and fancy indexing
- 3.5 Vectorised arithmetic and universal functions
- 3.6 Aggregation and axes
- 3.7 Broadcasting
- 3.8 Random numbers
- 3.9 Missing values and nan
- 3.10 A little linear algebra
- 3.11 Worked example: a week of sensor readings
- 3.12 Summary
- 3.13 Exercises
- 3.14 Quiz
Part II · Working with data
Chapter 4: pandas Fundamentals: Series, DataFrames and Selection
- 4.1 The Series
- 4.2 The DataFrame
- 4.3 Loading data
- 4.4 Inspecting a new table
- 4.5 Selecting columns and rows
- 4.6 Sorting and ranking
- 4.7 Adding and changing columns
- 4.8 The index
- 4.9 Modifying data safely: copy-on-write
- 4.10 Saving a table
- 4.11 Worked example: a first report on the orders
- 4.12 Summary
- 4.13 Exercises
- 4.14 Quiz
Chapter 5: Cleaning, Reshaping, Grouping and Joining
Chapter 6: Visualisation and Exploratory Data Analysis
- 6.1 Why look at the data?
- 6.2 How matplotlib works
- 6.3 A dataset to explore
- 6.4 One numerical variable: the distribution
- 6.5 Comparing groups
- 6.6 Relationships between two numerical variables
- 6.7 Change over time
- 6.8 Making charts honest and clear
- 6.9 An exploratory analysis, step by step
- 6.10 Summary
- 6.11 Exercises
- 6.12 Quiz
Part III · Statistics and models
Part IV · Further topics and practice
Back matter
About this book¶
A first course in data science with Python for readers who know school mathematics and a little programming. It follows the whole workflow: asking a question, NumPy arrays, loading, cleaning, reshaping, grouping and joining tables with pandas, exploratory analysis and plots with matplotlib, probability and statistical inference, regression, classification, model evaluation and cross-validation, clustering and dimensionality reduction, text and time series, and ethics, privacy and reproducibility, ending with a complete project. Every dataset is made in code and every program has been run, with its real output shown.