Leveraging OpenAI Gym and the Anytrading Environment for Trading~!pip install tensorflow-gpu==1.15.0 tensorflow==1.15.0 gym-anytrading gym stable-baselines Environment trading bot trade import gym_anytrading import gym from stable_baselines import A2C from stable_baselines.common.vec_env import DummyVevEnv import pandas as pd import numpy as np from matplotlib import pyplot as plt df=pd.read_csv(‘ETHUSD.csv’) df.head( )df[‘Date’]=pd.to_datetime(df[‘Date’])df.set_index(‘Date’,in place=True)df[‘Open’]=df[‘Open’].apply(lambda x:float(x.replace(“,”,””)))df[‘High’]=df[‘High’].apply(lambda x:float(x.replace(“,”,””)))df[‘Close’]=df[‘Close’].apply(lambda x:float(x.replace(“,”,””)))env=gym.make(‘stocks-v0’,df=df,frame_bound=(5,30),window_size=5)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=(20,10)) pl.cla( )env.render_all( ) pltl.show( ) env_build=lambda

:gym.make(‘atocks:

-vO’,df=df,frame_bound=(5,30),window_size=5)env=DummyVecEnv([env_build])model_train=A2C(‘MlpLstmPolicy’,env,verbose=1)model_train.learn(total_timesteps=100000)env=gym.make(‘stocks-v0’,df=df,frame_bound=(25,35),window_size=5)ob’s=env.reset( )while True:obs=obs[np.newaxis,…]action,_states=model_train.predict(obs)obs,rewards,done,info=dnv.step(action)if done:print(“info”,info)break plt.figure(figsize=(15,6))plt.cla( ) env.render_all( )plt.show( )