Python:股票×ETF量化交易實戰105個活用技巧

發表於 更新於 分類於 Disqus：
文章字數： 4.5k 所需閱讀時間 ≈ 4 分鐘

抓取公開資訊

Yahoo Finance

抓取每日歷史紀錄
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
# 抓取每日歷史紀錄
# pip3 install yfinance
import yfinance as yf

# 不含 end 那一天
# a= yf.download('0050.tw', start='2020-01-01', end='2024-09-30')
a= yf.download('0050.tw', start='2020-01-01', end='2024-10-01')
print(a)

#                   Open        High         Low       Close   Adj Close    Volume
# Date
# 2020-01-02   97.050003   98.000000   97.050003   97.650002   83.889481   4882015
# 2020-01-03   98.300003   98.699997   97.000000   97.650002   83.889481   6813547
# 2020-01-06   97.050003   97.099998   96.400002   96.400002   82.815636   9321768
# 2020-01-07   96.449997   96.699997   95.400002   96.099998   82.557899   6328602
# 2020-01-08   95.199997   96.199997   95.050003   95.650002   82.171310   8516995
# ...                ...         ...         ...         ...         ...       ...
# 2024-09-23  182.550003  183.649994  182.550003  183.250000  183.250000   7900498
# 2024-09-24  183.199997  185.000000  182.100006  185.000000  185.000000   9952346
# 2024-09-25  187.699997  188.000000  187.350006  187.750000  187.750000  11673001
# 2024-09-26  189.600006  190.000000  188.750000  189.300003  189.300003  16725931
# 2024-09-27  190.199997  191.250000  188.600006  188.899994  188.899994  11403314
# 2024-09-30  186.550003  187.199997  183.949997  183.949997  183.949997  13186950
# [1152 rows x 6 columns]
抓取5天,頻率為1分鐘
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
# 抓取5天,頻率為1分鐘
import yfinance as yf

a = yf.download('00632R.tw', period='5d', interval='1m')
print(a)
print(a.index[1])

#                           Open  High   Low  Close  Adj Close    Volume
# Datetime
# 2024-09-25 09:02:00+08:00  3.33  3.34  3.33   3.33       3.33         0
# 2024-09-25 09:03:00+08:00  3.33  3.34  3.33   3.33       3.33   7278000
# 2024-09-25 09:04:00+08:00  3.33  3.34  3.32   3.33       3.33  10622000
# 2024-09-25 09:05:00+08:00  3.33  3.34  3.32   3.33       3.33   5836000
# 2024-09-25 09:06:00+08:00  3.33  3.33  3.32   3.33       3.33  14689000
# ...                         ...   ...   ...    ...        ...       ...
# 2024-10-01 11:12:00+08:00  3.40  3.40  3.40   3.40       3.40     18000
# 2024-10-01 11:13:00+08:00  3.40  3.40  3.40   3.40       3.40    181000
# 2024-10-01 11:14:00+08:00  3.40  3.40  3.39   3.40       3.40    183000
# 2024-10-01 11:15:00+08:00  3.40  3.40  3.40   3.40       3.40     70000
# 2024-10-01 11:16:00+08:00  3.40  3.40  3.40   3.40       3.40         0

# [1186 rows x 6 columns]
# 2024-09-25 09:03:00+08:00
call getData()
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
# call getData()
from Data import getData
# 不含 end 那一天
# a = getData('0050', '2024-01-01', '2024-09-30')
a = getData('0050', '2024-01-02', '2024-10-01')
print(a)

#                   open        high         low       close   adj close    volume
# Date
# 2024-01-02  135.600006  135.949997  134.649994  134.899994  131.155762   5922076
# 2024-01-03  133.699997  133.899994  132.300003  132.550003  128.870987  13547475
# 2024-01-04  132.550003  132.750000  132.300003  132.500000  128.822372   4567593
# 2024-01-05  132.550003  132.949997  132.100006  132.149994  128.482086   3999671
# 2024-01-08  133.000000  133.600006  132.750000  132.750000  129.065430   9394720
# ...                ...         ...         ...         ...         ...       ...
# 2024-09-24  183.199997  185.000000  182.100006  185.000000  185.000000   9952346
# 2024-09-25  187.699997  188.000000  187.350006  187.750000  187.750000  11673001
# 2024-09-26  189.600006  190.000000  188.750000  189.300003  189.300003  16725931
# 2024-09-27  190.199997  191.250000  188.600006  188.899994  188.899994  11403314
# 2024-09-30  186.550003  187.199997  183.949997  183.949997  183.949997  13186950

# [180 rows x 6 columns]
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
# Data.py
import yfinance as yf
import pandas as pd
import os

datapath = 'data'

def getData(prod, st, en):
    if not os.path.exists(datapath):
        os.makedirs(datapath)

    bakfile = 'data//YF_%s_%s_%s_stock_daily_adj.csv' % (prod, st, en)
    if os.path.exists(bakfile):
        data = pd.read_csv(bakfile)
        data['Date'] = pd.to_datetime(data['Date'])
        data = data.set_index('Date')
    else:
        data = yf.download(f"{prod}.TW", start=st , end=en)
        data.columns = [i.lower() for i in data.columns]
        if data.shape[0] == 0:
            print('No data')
            return pd.DataFrame()
        data.to_csv(bakfile)
    return data

FinMind 套件

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
# package FinMind - 抓取每日歷史紀錄
# pip3 install FinMind
# https://www.twse.com.tw/zh/trading/historical/stock-day.html
from FinMind.data import DataLoader

FM = DataLoader()
a = FM.taiwan_stock_daily(stock_id='0050', start_date='2020-01-01', end_date='2024-10-01')
print(a)

#             date stock_id  Trading_Volume  Trading_money    open     max     min   close  spread  Trading_turnover
# 0     2020-01-02     0050         4882015      476649683   97.05   98.00   97.05   97.65    0.70              2421
# 1121  2024-09-25     0050        12464313     2339725240  187.70  188.00  187.35  197.65    0.00              3080
# 1121  2024-09-25     0050        12464313     2339725240  187.70  188.00  187.35  187.75    2.75             15933
# 1122  2024-09-26     0050        17369543     3290621957  189.60  190.00  188.75  189.30    1.55             15267
# 1123  2024-09-27     0050        11999510     2280157942  190.20  191.25  188.60  188.90   -0.40             15599
# 1124  2024-09-30     0050        14461065     2678744441  186.55  187.20  183.95  183.95   -4.95             32756
# 1125  2024-10-01     0050         9883064     1816503951  183.90  184.60  183.35  183.60   -0.35             12966
# [1126 rows x 10 columns]

Ref