0.3.0

[0.3.0] - 2023-12-05

Added:

• Added DifferentActions and AccountValue as metrics. Metrics are the main way to evaluate the performance of the agent.
• Now metrics.Metrics object can be used to calculate the metrics within trading environment.
• Included rockrl==0.0.4 as a dependency, which is a reinforcement learning package that I created.
• Added experiments/training_ppo_sinusoid.py to train a simple Dense agent using PPO algorithm on the sinusoid data with discrete actions.
• Added experiments/testing_ppo_sinusoid.py to test the trained agent on the sinusoid data with discrete actions.

Changed:

• Renamed and moved playing.py to experiments/playing_random_sinusoid.py
• Upgraded finrock.render.PygameRender, now we can stop/resume rendering with spacebar and render account value along with the actions