-
Notifications
You must be signed in to change notification settings - Fork 12
/
Copy pathch3.1to3.4_GetData.py
215 lines (190 loc) · 7.36 KB
/
ch3.1to3.4_GetData.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
##################################################
## CH 3.1. 使用yfinance取得日線資料
#################################################
import yfinance as yf
#告訴yfinance,要取0050的資料
tw0050 = yf.Ticker("0050.tw")
#取所有歷史資料
tw0050_max = tw0050.history(period="max")
print('tw0050_max:')
print(tw0050_max)
#取五年歷史資料
tw0050_5y = tw0050.history(period="5y")
print('tw0050_5y:')
print(tw0050_5y)
#取七天分的一分線歷史資料
tw0050_1m = tw0050.history(period="7d", interval = "1m")
#取收盤價
close=tw0050_max['Close']
#畫收盤價
import matplotlib.pyplot as plt
plt.plot(close)
plt.title('Get 0050 close with yfinance')
plt.show()
##################################################
## CH 3.2. 使用shioaji取得tick資料
#################################################
#登入Shioaji
import shioaji as sj
api = sj.Shioaji(simulation=False)
person_id=''#你的身分證字號
passwd=''#你的永豐證券登入密碼
if(person_id==''):
person_id=input("Please input ID:\n")
if(passwd==''):
passwd=input("Please input PASSWORD:\n")
api.login(
person_id=person_id,
passwd=passwd,
contracts_timeout=10000,
contracts_cb=lambda security_type: print(f"{repr(security_type)} fetch done.")
)
#登入永豐金證券的憑證
CA_passwd=''
if(CA_passwd==''):
CA_passwd=input("Please input CA PASSWORD:\n")
CA='c:\ekey\\551\\'+person_id+'\\S\\Sinopac.pfx'
result = api.activate_ca(\
ca_path=CA,\
ca_passwd=CA_passwd,\
person_id=person_id,\
)
#取得股票0050的物件
contract_0050 = api.Contracts.Stocks["0050"]
#取得小型台指近月的物件
contract_MXFR1 = api.Contracts.Futures['MXFR1']
#列出小型台指的不同月份的合約
print('api.Contracts.Futures.MXF content:')
print(api.Contracts.Futures.MXF)
#取得tick資料
ticks_0050 = api.ticks(
contract=contract_0050,
date="2021-12-15"
)
ticks_MXFR1 = api.ticks(
contract=contract_MXFR1,
date="2021-12-15"
)
#把0050的ticks轉成dataframe,並且印出資料
import pandas as pd
df_0050_tick = pd.DataFrame({**ticks_0050})#轉成dataframe
df_0050_tick.index =pd.to_datetime(df_0050_tick.ts)
#把小型台指近月的ticks轉成dataframe
df_MXFR1_tick = pd.DataFrame({**ticks_MXFR1})#轉成dataframe
df_MXFR1_tick.index = pd.to_datetime(df_MXFR1_tick.ts)
print('df_MXFR1_tick:')
print(df_MXFR1_tick)
plt.plot(df_MXFR1_tick['close'])
plt.title('Get MXFR1 ticks with shioaji')
plt.show()
#ticks轉1分K的範例程式
def ticksTo1mkbars(ticks):
period='1min'
kbars_out = pd.DataFrame(columns = ['Open','High','Low','Close','Volume'])
kbars_out['Open'] = ticks['close'].resample(period).first() #區間第一筆資料為開盤(Open)
kbars_out['High'] = ticks['close'].resample(period).max() #區間最大值為最高(High)
kbars_out['Low'] = ticks['close'].resample(period).min() #區間最小值為最低(Low)
kbars_out['Close'] = ticks['close'].resample(period).last() #區間最後一個值為收盤(Close)
kbars_out['Volume'] = ticks['volume'].resample(period).sum() #區間所有成交量加總
kbars_out=kbars_out.dropna()
return kbars_out
MXFR1_1m=ticksTo1mkbars(df_MXFR1_tick)
print('MXFR1_1m:')
print(MXFR1_1m)
plt.plot(MXFR1_1m['Close'])
plt.title('Convert to 1m data')
plt.show()
##################################################
## CH 3.3. 使用shioaji取得1分線資料
#################################################
#取得kbars資料
kbars_0050 = api.kbars(
contract_0050,
start="2021-12-01",
end="2021-12-30")
kbars_MXFR1 = api.kbars(
contract_MXFR1,
start="2021-12-01",
end="2021-12-30")
#把0050的kbars轉成dataframe,並且印出最前面的資料
import pandas as pd
df_0050 = pd.DataFrame({**kbars_0050})#轉成dataframe
df_0050.index =pd.to_datetime(df_0050.ts)
print('df_0050:')
print(df_0050)
#把小型台指近月的kbars轉成dataframe
df_MXFR1 = pd.DataFrame({**kbars_MXFR1})#轉成dataframe
df_MXFR1.index = pd.to_datetime(df_MXFR1.ts)
print('df_MXFR1:')
print(df_MXFR1)
plt.plot(df_MXFR1['Close'])
plt.title('Get 1m data with shioaji')
plt.show()
df_MXFR1_fix=df_MXFR1.groupby(df_MXFR1.index).first()
plt.plot(df_MXFR1_fix['Close'])
plt.title('Fix error in shioaji 1m data')
plt.show()
##################################################
## CH 3.4. 如何把1分線轉為小時線或其他週期
#################################################
#小型台指 1小時
period='1h'
kbars = pd.DataFrame(columns = ['Open','High','Low','Close','Volume'])
kbars['Open'] = df_MXFR1['Open'].resample(period).first() #區間第一筆資料為開盤(Open)
kbars['High'] = df_MXFR1['High'].resample(period).max() #區間最大值為最高(High)
kbars['Low'] = df_MXFR1['Low'].resample(period).min() #區間最小值為最低(Low)
kbars['Close'] = df_MXFR1['Close'].resample(period).last() #區間最後一個值為收盤(Close)
kbars['Volume'] = df_MXFR1['Volume'].resample(period).sum() #區間所有成交量加總
kbars_1h=kbars.dropna()
print('MXFR1 1h:')
print(kbars_1h)
plt.plot(kbars_1h['Close'])
plt.title('Resample MXFR1 to 1h')
plt.show()
#小型台指 日線
period='1d'
kbars = pd.DataFrame(columns = ['Open','High','Low','Close','Volume'])
kbars['Open'] = df_MXFR1['Open'].resample(period).first() #區間第一筆資料為開盤(Open)
kbars['High'] = df_MXFR1['High'].resample(period).max() #區間最大值為最高(High)
kbars['Low'] = df_MXFR1['Low'].resample(period).min() #區間最小值為最低(Low)
kbars['Close'] = df_MXFR1['Close'].resample(period).last() #區間最後一個值為收盤(Close)
kbars['Volume'] = df_MXFR1['Volume'].resample(period).sum() #區間所有成交量加總
kbars_1d=kbars.dropna()
print('MXFR1 1d:')
print(kbars_1d)
plt.plot(kbars_1d['Close'])
plt.title('Resample MXFR1 to 1d')
plt.show()
#小型台指 15分線
period='15min'
kbars = pd.DataFrame(columns = ['Open','High','Low','Close','Volume'])
kbars['Open'] = df_MXFR1['Open'].resample(period).first() #區間第一筆資料為開盤(Open)
kbars['High'] = df_MXFR1['High'].resample(period).max() #區間最大值為最高(High)
kbars['Low'] = df_MXFR1['Low'].resample(period).min() #區間最小值為最低(Low)
kbars['Close'] = df_MXFR1['Close'].resample(period).last() #區間最後一個值為收盤(Close)
kbars['Volume'] = df_MXFR1['Volume'].resample(period).sum() #區間所有成交量加總
kbars_15m=kbars.dropna()
print('MXFR1 15m:')
print(kbars_15m)
plt.plot(kbars_15m['Close'])
plt.title('Resample MXFR1 to 15m')
plt.show()
#寫成函數
def sjBarsToDf(sjBars):
dfBars = pd.DataFrame({**sjBars})#轉成dataframe
dfBars.index =pd.to_datetime(dfBars.ts)
dfBars=dfBars.groupby(dfBars.index).first()
return dfBars
def resampleKbars(kbars,period='1d'):
kbars_out = pd.DataFrame(columns = ['Open','High','Low','Close','Volume'])
kbars_out['Open'] = kbars['Open'].resample(period).first() #區間第一筆資料為開盤(Open)
kbars_out['High'] = kbars['High'].resample(period).max() #區間最大值為最高(High)
kbars_out['Low'] = kbars['Low'].resample(period).min() #區間最小值為最低(Low)
kbars_out['Close'] = kbars['Close'].resample(period).last() #區間最後一個值為收盤(Close)
kbars_out['Volume'] = kbars['Volume'].resample(period).sum() #區間所有成交量加總
kbars_out=kbars_out.dropna()
return kbars_out
df=sjBarsToDf(kbars_MXFR1)
df_resample=resampleKbars(df,period='1d')
print('MXFR1 1d (generated with function):')
print(df_resample)