依據文檔建議, 入門先看 Stock_NeurIPS2018 系列教學
參考資料 Stock_NeurIPS2018_1_Data.ipynb

直接透過FinRL API的YahooDownloader抓數據

• FinRL格式: 用 Adjust Close 取代 Close
  FinRL抓取資料後會調整欄位，close的值其實是原始資料中的adjust close，而原始的close資料會被拋棄
• date 與 tic 欄位:
  • date: 原始數據中的datetime
  • tic: 股票代號

         date       open       high        low      close     volume   tic  day
0  2020-01-02  74.059998  75.150002  73.797501  72.876114  135480400  aapl    3
1  2020-01-03  74.287498  75.144997  74.125000  72.167610  146322800  aapl    4
2  2020-01-06  73.447502  74.989998  73.187500  72.742683  118387200  aapl    0

import pandas as pd
from finrl.meta.preprocessor.yahoodownloader import YahooDownloader
from finrl.meta.preprocessor.preprocessors import FeatureEngineer, data_split
from finrl import config_tickers
from finrl.config import INDICATORS

import itertools

aapl_df_finrl = YahooDownloader(start_date = '2020-01-01',
                                end_date = '2020-01-31',
                                ticker_list = ['aapl']).fetch_data()
aapl_df_finrl.head()

TRAIN_START_DATE = '2009-01-01'
TRAIN_END_DATE = '2020-07-01'
TRADE_START_DATE = '2020-07-01'
TRADE_END_DATE = '2021-10-29'
df_raw = YahooDownloader(start_date = TRAIN_START_DATE,
                     end_date = TRADE_END_DATE,
                     ticker_list = config_tickers.DOW_30_TICKER).fetch_data()
df_raw.head()

fe = FeatureEngineer(use_technical_indicator=True,
                     tech_indicator_list = INDICATORS,
                     use_vix=True,
                     use_turbulence=True,
                     user_defined_feature = False)
processed = fe.preprocess_data(df_raw)
list_ticker = processed["tic"].unique().tolist()
list_date = list(pd.date_range(processed['date'].min(),processed['date'].max()).astype(str))
combination = list(itertools.product(list_date,list_ticker))

processed_full = pd.DataFrame(combination,columns=["date","tic"]).merge(processed,on=["date","tic"],how="left")
processed_full = processed_full[processed_full['date'].isin(processed['date'])]
processed_full = processed_full.sort_values(['date','tic'])

processed_full = processed_full.fillna(0)
processed_full.head()

資料預處理後, 直接可以作為之後訓練與預測的輸入

train = data_split(processed_full, TRAIN_START_DATE,TRAIN_END_DATE)
trade = data_split(processed_full, TRADE_START_DATE,TRADE_END_DATE)
print(len(train))
print(len(trade))

train.to_csv('train_data.csv')
trade.to_csv('trade_data.csv')