Cheat Sheets

Dense, one-page references to keep open while you work. They summarize what the chapters teach in depth. If something on a cheat sheet doesn't make sense, follow it back to its chapter.

Data and math

• NumPy

  ---

  Array creation, indexing, broadcasting, reductions, linear algebra, and random numbers.

• pandas

  ---

  Selection, cleaning, groupby, joins, reshaping, time series, and method chaining.

• Visualization

  ---

  Matplotlib and seaborn recipes, and which chart to use when.

• SQL

  ---

  Queries, joins, CTEs, and window functions for analysis.

• Math notation

  ---

  Every symbol in the handbook, Greek letters, and the key formulas.

• Probability and statistics

  ---

  Distributions, which test to use, and the essential formulas.

Machine learning

• scikit-learn

  ---

  The estimator API, preprocessing, models, model selection, and pipelines.

• Choosing a model

  ---

  A decision flowchart and a comparison of every algorithm in the handbook.

• Evaluation metrics

  ---

  Every metric, with its formula, when to use it, and its scikit-learn function.

Deep learning

• PyTorch

  ---

  Tensors, autograd, modules, data loading, and the training loop.

• Neural network recipes

  ---

  Losses, activations, initializations, optimizer defaults, shape cheats, and a debugging checklist.

• Transformers and LLMs

  ---

  Attention formulas, block anatomy, parameter counting, sampling, fine-tuning, and RAG.

Reference

• Glossary

  ---

  Every term in the handbook, defined in one place.