How to build stock portfolio using deep reinforcement learning in Python by taking care of indicators for each given stock?

External question — Data Science Stack Exchange Author: YoYoYo Original post: https://datascience.stackexchange.com/questions/115167 License: CC BY-SA 4.0 — https://creativecommons.org/licenses/by-sa/4.0/ Adaptation: HTML converted to plain text; contact email addresses removed. I have this Python code from this tutorial which is trading a stock (e.g. GME but it can be any) by taking care of its indicators ('SMA', 'RSI', 'OBV'). # Gym stuff import gym import gym_anytrading # Stable baselines - rl stuff from stable_baselines.common.vec_env import DummyVecEnv from stable_baselines import A2C # Processing libraries import numpy as np import pandas as pd from matplotlib import pyplot as plt # Custom indicators libraries: from gym_anytrading.envs import StocksEnv from finta import TA df = pd.read_csv('data/gmedata.csv') df['Date'] = pd.to_datetime(df['Date']) df.sort_values('Date', ascending=True, inplace=True) df.set_index('Date', inplace=True) env = gym.make('stocks-v0', df=df, frame_bound=(5,250), window_size=5) # Build Environment: state = env.reset() while True: action = env.action_space.sample() n_state, reward, done, info = env.step(action) if done: print("info", info) break #plt.figure(figsize=(15,6)) #plt.cla() #env.render_all() #plt.show() # Fix Volume Column: df['Volume'] = df['Volume'].apply(lambda x: float(x.replace(",", ""))) # Calculate SMA, RSI and OBV: df['SMA'] = TA.SMA(df, 12) df['RSI'] = TA.RSI(df) df['OBV'] = TA.OBV(df) df.fillna(0, inplace=True) # Create New Environments: def add_signals(env): start = env.frame_bound[0] - env.window_size end = env.frame_bound[1] prices = env.df.loc[:, 'Low'].to_numpy()[start:end] signal_features = env.df.loc[:, ['Low', 'Volume','SMA', 'RSI', 'OBV']].to_numpy()[start:end] return prices, signal_features class MyCustomEnv(StocksEnv): _process_data = add_signals env2 = MyCustomEnv(df=df, window_size=12, frame_bound=(12,50)) # Build Environment and Train: env_maker = lambda: env2 env = DummyVecEnv([env_maker]) model = A2C('MlpLstmPolicy', env, verbose=1) model.learn(total_timesteps=1000000) # Evaluation: env = MyCustomEnv(df=df, window_size=12, frame_bound=(80,250)) obs = env.reset() while True: obs = obs[np.newaxis, ...] action, _states = model.predict(obs) obs, rewards, done, info = env.step(action) if done: print("info", info) break plt.figure(figsize=(15,6)) plt.cla() env.render_all() plt.show() Here is the github repo for this code where you can find also the GME stock dataset too: https://github.com/nicknochnack/Reinforcement-Learning-for-Trading-Custom-Signals/blob/main/Custom%20Signals.ipynb Now, let's say instead of feeding this deep RL code with just one stock ticker (GME) with its indicators values ('SMA', 'RSI', 'OBV') you feed it with a bunch of tickers (e.g. 100 or 1000 or so) and each of them to have their own indicators values. How such a dataset should look like and how this code should be in order to accomplish such a thing because currently it is working for just one single ticker?