Python 101 — Basics for a New Bee
How do we turn CSV records into a useful command-line summary?
python python_101.py python python_101.py --self-test
6 checks passed · Python standard library
Working examples you can run.
Each example includes imports and generated data. Working code and tests are included. No program imports another lesson's example.
Download the legacy coding bundle · Setup and command reference
Sixteen programs use only Python's standard library. Five ML programs require PyTorch; PINN also uses NumPy and Matplotlib. All were executed from isolated temporary directories.
Run your first example
python3 -m venv .venv source .venv/bin/activate python python_101.py python python_101.py --self-test # For ML examples: python -m pip install -r requirements.txt python cnn.py python cnn.py --self-test
Use Python 3.10+. On Windows PowerShell, activate with .venv\Scripts\Activate.ps1. ML demos run on CPU; no network dataset download is needed.
Programming foundations
The Science of Programming — Explained Simply
How can a contract determine the loop we should write?
python science_of_programming.py python science_of_programming.py --self-test
660 checks passed · Python standard library
Weakest Preconditions — Explained Simply
Which starting states let a program finish with the required result?
python weakest_preconditions.py python weakest_preconditions.py --self-test
533 checks passed · Python standard library
The Science of Programming — A Practical Python Playbook
How do we implement and check loops with precise boundary behavior?
python science_of_programming_python.py python science_of_programming_python.py --self-test
2304 checks passed · Python standard library
Neural networks and learning
PyTorch 101 — From Tensors to Training
Can gradients learn the unknown slope and intercept of a line?
python pytorch_101.py python pytorch_101.py --self-test
5 checks passed · torch>=2.2
Neural Networks and CNNs — From First Principles
Can shared local filters learn to distinguish vertical and horizontal bars?
python cnn.py python cnn.py --self-test
6 checks passed · torch>=2.2
How LLMs Work — A Tech Talk
Can a small model learn next-character prediction while being unable to see future targets?
python llm_tech_talk.py python llm_tech_talk.py --self-test
5 checks passed · torch>=2.2
Adaptive Neural Infrastructure with Memory Accumulation
Can we select experts, accumulate state, and keep memory scoped to a user?
python adaptive_neural_infra.py python adaptive_neural_infra.py --self-test
8 checks passed · torch>=2.2
Physics-Informed Neural Networks
Train a heat-equation PINN, inspect its loss components, and compare against the exact solution.
python pinn_heat_equation.py --epochs 2500 python pinn_heat_equation.py --self-test
4 checks passed · torch, numpy, matplotlib
PINN for software engineers
Tokens, rules, and a real PINN
Use 192 generated Python programs for next-token training. Compare a grammar penalty with an inference mask.
The second lab learns a decay equation without sensor records.
Computation and algorithms
Automata Theory — DFA, NFA & Regular Expressions
Can a small state machine recognize a pattern one symbol at a time?
python automata_theory.py python automata_theory.py --self-test
510 checks passed · Python standard library
Context-Free Grammars & Pushdown Automata
How does a stack or parsing table recognize nested structure?
python cfg_pda.py python cfg_pda.py --self-test
254 checks passed · Python standard library
Turing Machines & Undecidability
How can a head and a transition table transform a binary number?
python turing_machines.py python turing_machines.py --self-test
257 checks passed · Python standard library
P vs NP & NP-Completeness
How do we check a SAT witness and preserve satisfiability in a graph reduction?
python p_vs_np.py python p_vs_np.py --self-test
42 checks passed · Python standard library
Lambda Calculus — Computation from Functions Alone
Can function application compute while preserving which binder a variable belongs to?
python lambda_calculus.py python lambda_calculus.py --self-test
6 checks passed · Python standard library
Randomized Algorithms — Probability as a Tool
Which guarantees come from correctness, and which depend on random choices?
python randomized_algorithms.py python randomized_algorithms.py --self-test
389 checks passed · Python standard library
Information and compression
Entropy & Shannon's Source Coding Theorem
Can a prefix code reduce the average bits needed for a skewed source?
python entropy_source_coding.py python entropy_source_coding.py --self-test
14 checks passed · Python standard library
Channel Capacity & Error-Correcting Codes
How does structured redundancy correct a single flipped bit?
python channel_capacity_codes.py python channel_capacity_codes.py --self-test
131 checks passed · Python standard library
KL Divergence, Cross-Entropy & Machine Learning
How do model probabilities turn into a stable loss and gradient?
python kl_cross_entropy.py python kl_cross_entropy.py --self-test
14 checks passed · Python standard library
Kolmogorov Complexity — The Information in One Object
What can a short description show that a compressor's output length cannot?
python kolmogorov_complexity.py python kolmogorov_complexity.py --self-test
1005 checks passed · Python standard library
Compression in Practice — Arithmetic Coding, LZ & Beyond
Can we encode data compactly and decode it exactly?
python compression_in_practice.py python compression_in_practice.py --self-test
29 checks passed · Python standard library
Functional composition
Monads in Python — With the Math
How can we compose successful steps while keeping missing values or errors explicit?
python monads_python.py python monads_python.py --self-test
38 checks passed · Python standard library
Captured outputs and numerical results are in verification.json. Tests cover finite cases and complement mathematical reasoning; they are not universal proofs.