TradeMaster-NTU/TradeMaster — reverse-engineered prompt

Build me a research platform for quantitative trading using reinforcement learning, like a complete workspace where I can load market datasets, train trading agents, test them, compare results, and visualize how they perform.

I want it to feel usable for someone learning this area, with clear tutorials or notebooks, sample stock and futures data, example strategies, and a simple way to run experiments without having to wire everything together manually. Please include the full flow from data preparation to training, evaluation, charts, and saving models so they can be reused later.

It should support common trading research tasks like portfolio management, market simulation, technical indicators, feature generation, and comparing different reinforcement learning methods. Please keep it organized, documented, and runnable on a normal Python setup. If you need current docs for reinforcement learning trading tools, look them up online. Don’t make it trade real money by default, this should be for research and experimentation first.