How khichinho/mdp-blackjack works
Project type and purpose
• This is an algorithmic/open-source tool (not a full-stack app) for modeling and solving a Blackjack game as a Markov Decision Process (MDP). The README title “MDP-Blackjack” and filenames like Markov.py and Iteration.py point to a simulation/solver framework.
File structure
• Top-level Python modules:
• Dealer.py – simulates the dealer (environment dynamics)
• Markov.py – defines states, actions, transition probabilities, rewards
• Iteration.py – implements value or policy iteration over the MDP
• FormatCheck.py – utility to validate input/output data formats
• run.sh – shell script entry point that ties everything together
• .gitignore / LICENSE – project metadata
• No frontend or external services, purely a CLI‐driven MDP solver.
Detailed component breakdown
• “FormatCheck.py” (validates inputs)
• “Markov.py” → supplies transitions & rewards
• “Dealer.py” → supplies environment rollouts
• “Iteration.py” → consumes data from Markov.py and Dealer.py, outputs policy/value files