Pandas 策略開發基礎語法

1.安裝環境¶

In [ ]:

!pip install finlab > log.txt

!pip install finlab > log.txt

In [ ]:

import pandas as pd
import numpy as np
from finlab import data

import pandas as pd import numpy as np from finlab import data

2.創造資料物件¶

In [ ]:

# dataframe:資料格式範例
df=data.get('price:收盤價')
df=df[df.index<='2020-07-01']

# series:資料格式範例
series=df['0050']
df

# dataframe:資料格式範例 df=data.get('price:收盤價') df=df[df.index<='2020-07-01'] # series:資料格式範例 series=df['0050'] df

3.資料檢索¶

In [ ]:

df=data.get('price:收盤價')

# head(n):取前n列，預設為5
demo_head=df.head(3)

# tail(n):取後n列，預設為5
demo_tail=df.tail(3)

# index:取索引
demo_index=df.index

# columns:取欄位
demo_columns=df.columns

# values:取值
demo_values=df.values

# boolean indexing:索引布林邏輯篩選，ex:取得索引大於2020年份的資料
demo_boolean_indexing=df[df.index>'2020']

# loc:使用標籤篩選欄列
## ex:選擇2020-01-03~2020-01-08中所有標地的股價
demo_loc1=df.loc['2020-01-03':'2020-01-08']

## ex:選擇0050&1101&2330的股價，以下兩種方法等價
demo_loc2=df.loc[:,['0050','1101','2330']]
demo_column_select=df[['0050','1101','2330']]

## ex:選擇2020-01-03~2020-01-08中0050&1101&2330的股價
demo_loc3=df.loc['2020-01-03':'2020-01-08',['0050','1101','2330']]

# iloc:使用標籤位置篩選欄列
## ex:選擇第6列到第10列的資料
demo_iloc1=df.iloc[5:10]

## ex:選擇df表中第6檔到第10檔標地資料
demo_iloc2=df.iloc[:,5:10]

## ex:選擇第6列到第10列的資料&選擇df表中第6檔到第10檔標地資料
demo_iloc3=df.iloc[5:10,5:10]

# nlargest:取前n大，ex:取近一日股價前n大的標地
demo_nlargest=df.iloc[-1].nlargest(10)

# nsmallest:取前n大，ex:取近一日股價前n小的標地
demo_nsmallest=df.iloc[-1].nsmallest(10)

df=data.get('price:收盤價') # head(n):取前n列，預設為5 demo_head=df.head(3) # tail(n):取後n列，預設為5 demo_tail=df.tail(3) # index:取索引 demo_index=df.index # columns:取欄位 demo_columns=df.columns # values:取值 demo_values=df.values # boolean indexing:索引布林邏輯篩選，ex:取得索引大於2020年份的資料 demo_boolean_indexing=df[df.index>'2020'] # loc:使用標籤篩選欄列 ## ex:選擇2020-01-03~2020-01-08中所有標地的股價 demo_loc1=df.loc['2020-01-03':'2020-01-08'] ## ex:選擇0050&1101&2330的股價，以下兩種方法等價 demo_loc2=df.loc[:,['0050','1101','2330']] demo_column_select=df[['0050','1101','2330']] ## ex:選擇2020-01-03~2020-01-08中0050&1101&2330的股價 demo_loc3=df.loc['2020-01-03':'2020-01-08',['0050','1101','2330']] # iloc:使用標籤位置篩選欄列 ## ex:選擇第6列到第10列的資料 demo_iloc1=df.iloc[5:10] ## ex:選擇df表中第6檔到第10檔標地資料 demo_iloc2=df.iloc[:,5:10] ## ex:選擇第6列到第10列的資料&選擇df表中第6檔到第10檔標地資料 demo_iloc3=df.iloc[5:10,5:10] # nlargest:取前n大，ex:取近一日股價前n大的標地 demo_nlargest=df.iloc[-1].nlargest(10) # nsmallest:取前n大，ex:取近一日股價前n小的標地 demo_nsmallest=df.iloc[-1].nsmallest(10)

4.顯示控制¶

pandas預防資源消耗，預設顯示列數為10，中間資料以‘...’帶過，若想要全展開或調整顯示數量來檢視資料，可作以下控制，要注意的是此影響為全域，接下來的顯示數都會被影響。

In [ ]:

df=data.get('price:收盤價')

# 列數全展開
pd.set_option("display.max_rows", None)

# 欄數全展開
pd.set_option("display.max_columns", None)

# 最多顯示20列
pd.set_option("display.max_rows", 20)

# 最多顯示20欄
pd.set_option("display.max_columns", 20)

# # 還原顯示數預設初始值
pd.reset_option("^display")
df

df=data.get('price:收盤價') # 列數全展開 pd.set_option("display.max_rows", None) # 欄數全展開 pd.set_option("display.max_columns", None) # 最多顯示20列 pd.set_option("display.max_rows", 20) # 最多顯示20欄 pd.set_option("display.max_columns", 20) # # 還原顯示數預設初始值 pd.reset_option("^display") df

Out[ ]:

	0015	0050	0051	0052	0053	0054	0055	0056	0057	0058	0059	0060	0061	006201	006202	006203	006204	006205	006206	006207	006208	00625K	00631L	00632R	00633L	00634R	00635U	00636	00636K	00637L	00638R	00639	00640L	00641R	00642U	00643	00643K	00645	00646	00647L	...	9908	9910	9911	9912	9914	9915	9917	9918	9919	9921	9922	9924	9925	9926	9927	9928	9929	9930	9931	9933	9934	9935	9937	9938	9939	9940	9941	9941A	9942	9943	9944	9945	9946	9949	9950	9951	9955	9958	9960	9962
date
2007-04-23	9.54	57.85	32.83	38.40	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	18.60	28.00	10.35	14.00	48.55	32.35	56.5	21.40	16.55	52.0	1.87	29.35	31.05	16.60	34.60	18.30	4.80	37.00	15.80	28.55	48.55	9.93	27.55	31.25	30.80	95.00	21.60	NaN	48.70	19.70	31.00	30.35	9.78	73.60	13.25	51.3	72.40	NaN	46.00	49.60
2007-04-24	9.54	58.10	32.99	38.65	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	18.50	27.75	10.35	14.40	48.95	32.60	55.5	21.30	16.40	52.5	1.83	29.30	30.95	16.45	34.80	18.25	4.82	39.55	15.30	28.30	47.80	9.75	27.30	31.40	30.60	94.50	21.50	NaN	48.70	19.65	32.00	30.50	9.91	75.00	13.25	50.5	71.60	NaN	45.90	50.40
2007-04-25	9.52	57.60	32.80	38.59	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	18.10	27.80	10.20	14.00	49.75	32.10	55.5	21.20	17.35	54.1	1.83	29.15	30.70	16.40	34.95	18.15	4.71	42.30	16.20	28.10	46.80	9.68	28.35	31.30	30.00	94.40	21.55	NaN	48.60	19.65	32.30	29.50	10.10	74.60	13.30	49.9	71.60	NaN	49.10	49.10
2007-04-26	9.59	57.70	32.80	38.60	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	18.25	27.90	10.05	14.00	49.60	32.50	57.4	21.05	16.95	55.4	1.90	29.20	30.70	16.40	34.35	18.05	4.81	42.30	15.80	28.00	46.80	9.81	28.20	31.45	29.75	94.90	21.65	NaN	47.65	19.50	31.60	29.15	10.80	74.50	13.25	49.5	71.00	NaN	48.90	48.00
2007-04-27	9.55	57.50	32.72	38.40	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	...	18.25	27.75	9.90	14.00	49.00	31.65	56.8	21.00	17.20	55.6	1.77	29.35	30.70	16.60	34.05	18.00	4.85	40.00	15.70	27.80	44.70	10.20	28.20	29.50	29.35	95.00	21.60	NaN	47.80	19.35	31.40	28.50	11.55	75.70	13.15	48.8	69.50	NaN	48.60	46.50
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
2021-06-24	NaN	136.70	57.60	124.70	65.75	NaN	21.17	34.66	93.75	NaN	NaN	NaN	23.55	20.76	NaN	65.20	89.95	38.10	38.00	31.06	78.65	NaN	127.75	5.62	67.30	3.86	23.82	27.22	9.71	26.55	7.77	17.05	28.49	10.62	12.50	18.51	NaN	25.48	36.48	63.80	...	33.40	223.00	67.00	8.93	312.00	NaN	93.6	40.20	34.60	319.0	NaN	44.60	38.65	48.10	35.40	5.41	11.40	47.00	39.90	38.75	17.05	21.65	47.75	100.50	76.00	30.60	138.50	51.2	96.20	64.00	22.00	54.20	14.80	10.80	12.35	94.5	18.75	134.5	27.55	11.00
2021-06-25	NaN	136.95	57.85	124.30	66.30	31.84	21.25	34.74	94.05	NaN	NaN	NaN	23.94	20.86	NaN	65.50	90.50	38.65	38.55	31.49	78.90	NaN	128.65	5.61	69.65	3.79	23.82	27.73	9.93	27.54	7.64	17.33	28.91	10.52	12.51	18.77	NaN	25.67	36.54	64.35	...	33.45	241.50	66.00	9.00	320.00	NaN	93.8	39.90	34.60	322.5	NaN	45.15	38.70	48.25	37.50	5.50	11.40	47.00	39.70	38.35	17.15	22.25	48.20	101.00	76.00	30.20	138.00	51.2	95.30	63.80	22.25	54.80	15.00	10.80	12.30	94.1	18.85	131.5	27.55	12.10
2021-06-28	NaN	137.20	57.90	124.20	66.15	31.84	21.29	35.10	94.05	NaN	NaN	NaN	24.00	20.83	NaN	65.45	91.10	38.53	38.40	31.63	79.20	NaN	130.30	5.57	69.15	3.80	23.92	27.63	9.90	27.35	7.59	17.55	28.96	10.51	12.64	18.98	NaN	25.67	36.68	64.70	...	33.45	238.50	66.70	9.39	317.00	NaN	94.4	39.90	34.70	325.0	NaN	45.35	38.85	48.25	37.40	5.45	11.55	47.40	39.70	38.50	17.20	23.30	47.50	103.00	77.00	30.00	138.00	51.3	99.40	63.20	22.15	57.60	14.80	10.80	12.30	95.5	19.00	131.5	27.10	13.30
2021-06-29	NaN	137.60	57.80	124.65	66.45	31.80	21.20	34.80	94.70	NaN	NaN	NaN	23.86	20.76	NaN	66.00	91.45	38.37	38.12	31.30	79.30	8.98	130.10	5.58	67.95	3.82	23.81	27.44	9.80	26.87	7.70	17.43	28.57	10.62	12.42	18.89	NaN	25.48	36.61	64.80	...	33.50	233.50	66.50	9.09	314.00	NaN	94.2	40.05	34.25	322.0	NaN	45.45	38.85	47.70	37.20	5.60	11.40	47.25	39.50	39.10	17.15	22.35	47.90	100.00	76.60	29.65	138.00	51.3	98.40	63.00	22.15	56.70	14.75	10.80	12.40	94.1	19.15	131.5	27.20	14.60
2021-06-30	NaN	138.95	58.20	125.00	66.80	31.90	21.26	35.00	95.45	NaN	NaN	NaN	23.86	20.98	NaN	66.30	92.60	38.40	38.30	31.45	79.90	NaN	133.05	5.50	68.20	3.81	23.60	27.45	9.83	26.96	7.65	17.55	28.62	10.58	12.54	18.92	NaN	25.51	36.72	65.00	...	33.55	244.50	66.40	9.10	313.00	NaN	94.1	40.10	34.70	318.5	NaN	45.85	38.90	47.80	38.25	5.50	11.40	47.50	39.75	39.20	17.30	22.40	48.00	102.00	76.90	29.60	142.50	51.3	108.00	63.50	22.15	56.60	14.70	10.85	12.55	93.8	19.30	134.0	27.35	16.05

3502 rows × 2268 columns

5.資料運算¶

In [ ]:

df=data.get('price:收盤價')

# 產生股價大於100元的布林dataframe，若大於100則顯示True(計算時視為1)，若不大於100則顯示False(計算時視為0)
price_up_10=df>100

# sum:加總，axis=0為整欄加總，axis=1為整列加總。ex:計算每日股價大於100的標地數量
demo_sum=price_up_10.sum(axis=1)

# mean:平均數，常用於計算均線。
demo_mean=df.mean()

# median:中位數
demo_median=df.median()

# std:標準差，常用於計算乖離率。
demo_std=df.std()

# max:最大值
demo_max=df.max()

# min:最小值
demo_min=df.min()

# cumsum:累加
## ex:累積漲多少元
demo_diff=df.diff()
demo_cumsum=demo_diff.cumsum()

# cumprod:累乘
## ex:計算累積報酬率
demo_pct_change=df.pct_change()+1
demo_cumprod=demo_pct_change.cumprod()

# cummax:累積最大值，常用於計算drawdown
demo_cummax=df.cummax()

# cummin:累積最小值
demo_cummin=df.cummin()

# quantile:第c百分位數數值，常用於取標的數前c%強標地
demo_quantile=df.iloc[-1].quantile(0.9)

# corr:相關性
demo_corr=df.iloc[:,:5].corr()

# describe:取統計資料
demo_describe=df.describe()

df=data.get('price:收盤價') # 產生股價大於100元的布林dataframe，若大於100則顯示True(計算時視為1)，若不大於100則顯示False(計算時視為0) price_up_10=df>100 # sum:加總，axis=0為整欄加總，axis=1為整列加總。ex:計算每日股價大於100的標地數量 demo_sum=price_up_10.sum(axis=1) # mean:平均數，常用於計算均線。 demo_mean=df.mean() # median:中位數 demo_median=df.median() # std:標準差，常用於計算乖離率。 demo_std=df.std() # max:最大值 demo_max=df.max() # min:最小值 demo_min=df.min() # cumsum:累加 ## ex:累積漲多少元 demo_diff=df.diff() demo_cumsum=demo_diff.cumsum() # cumprod:累乘 ## ex:計算累積報酬率 demo_pct_change=df.pct_change()+1 demo_cumprod=demo_pct_change.cumprod() # cummax:累積最大值，常用於計算drawdown demo_cummax=df.cummax() # cummin:累積最小值 demo_cummin=df.cummin() # quantile:第c百分位數數值，常用於取標的數前c%強標地 demo_quantile=df.iloc[-1].quantile(0.9) # corr:相關性 demo_corr=df.iloc[:,:5].corr() # describe:取統計資料 demo_describe=df.describe()

6.移動窗格作業¶

In [ ]:

df=data.get('price:收盤價')

# shift:資料移動，常用於前後期數值增減比較、年增率計算
## ex:收盤價向下平移一列
demo_shift1=df.shift()
## ex:收盤價向上平移一列
demo_shift2=df.shift(-1)

# rolling:移動窗格作業，常結合資料運算公式做滾動式計算，ex:計算20日移動平均線(前n筆數資料未滿10日取na，未滿20日以n日計算)
demo_rolling=df.rolling(20,min_periods=10).mean()

# diff:列數相減，ex:取每日漲跌價
demo_diff=df.diff()

# pct_change:列數相除，ex:取每日漲跌幅
demo_pct_change=df.pct_change()

df=data.get('price:收盤價') # shift:資料移動，常用於前後期數值增減比較、年增率計算 ## ex:收盤價向下平移一列 demo_shift1=df.shift() ## ex:收盤價向上平移一列 demo_shift2=df.shift(-1) # rolling:移動窗格作業，常結合資料運算公式做滾動式計算，ex:計算20日移動平均線(前n筆數資料未滿10日取na，未滿20日以n日計算) demo_rolling=df.rolling(20,min_periods=10).mean() # diff:列數相減，ex:取每日漲跌價 demo_diff=df.diff() # pct_change:列數相除，ex:取每日漲跌幅 demo_pct_change=df.pct_change()

7.使用FinlabDataFrame的功能¶

除了使用熟悉的 Pandas 語法，還可以使用FinlabDataFrame的function!

data.get()的資料繼承FinlabDataFrame的功能，function下列:

1. average(n):取移動平均，min_periods=int(n/2)。
2. is_largest(n):取前n大。
3. is_smallest(n):取前n小。
4. rise(n=1):取比前第n筆高。
5. fall(n=1):取比前第n筆低。
6. sustain(nwindow, nsatisfy=None):取滾動nwindow筆加總大於nsatisfy。
7. quantile_row(n):返回請求軸上給n定分位數的值。

In [ ]:

df[['0050','1101']].average(10)
df.iloc[-10:].is_largest(5)
df.iloc[-10:].is_smallest(5)
df.iloc[-10:,10:15].rise()
df.iloc[-10:,10:15].fall()
# ex:是否連兩日上漲？
df.iloc[-10:,10:15].rise().sustain(2)
# ex:取近10日股價前90％分位數
df.iloc[-10:].quantile_row(0.9)

df[['0050','1101']].average(10) df.iloc[-10:].is_largest(5) df.iloc[-10:].is_smallest(5) df.iloc[-10:,10:15].rise() df.iloc[-10:,10:15].fall() # ex:是否連兩日上漲？ df.iloc[-10:,10:15].rise().sustain(2) # ex:取近10日股價前90％分位數 df.iloc[-10:].quantile_row(0.9)

8.視覺化¶

8-1.乖離率線圖¶

In [ ]:

df=data.get('price:收盤價').iloc[-400:]['2330']
mean_20=df.rolling(20).mean()
std_value=df.rolling(20).std()

up=mean_20+std_value*2
down=mean_20-std_value*2
up.plot(label='up',legend=True)
down.plot(label='down',legend=True)
mean_20.plot(label='ma_20',legend=True)
df.plot(title='bias plot',label='close',legend=True,figsize=(20, 8),grid=True)

df=data.get('price:收盤價').iloc[-400:]['2330'] mean_20=df.rolling(20).mean() std_value=df.rolling(20).std() up=mean_20+std_value*2 down=mean_20-std_value*2 up.plot(label='up',legend=True) down.plot(label='down',legend=True) mean_20.plot(label='ma_20',legend=True) df.plot(title='bias plot',label='close',legend=True,figsize=(20, 8),grid=True)

Out[ ]:

<matplotlib.axes._subplots.AxesSubplot at 0x7f1ba253d350>

8-2.相關性熱力圖¶

In [ ]:

df=data.get('price:收盤價')
check_list=['1101','1102','2330','2454','6263','9939']
demo_corr=df.iloc[-600:][check_list].corr()

demo_corr.style.background_gradient(cmap ='viridis')\
        .set_properties(**{'font-size': '20px'})

df=data.get('price:收盤價') check_list=['1101','1102','2330','2454','6263','9939'] demo_corr=df.iloc[-600:][check_list].corr() demo_corr.style.background_gradient(cmap ='viridis')\ .set_properties(**{'font-size': '20px'})

Out[ ]:

	1101	1102	2330	2454	6263	9939
1101	1.000000	0.836385	0.562459	0.636151	0.571586	0.766617
1102	0.836385	1.000000	0.383488	0.445629	0.627689	0.754899
2330	0.562459	0.383488	1.000000	0.976506	0.093435	0.777646
2454	0.636151	0.445629	0.976506	1.000000	0.204041	0.808426
6263	0.571586	0.627689	0.093435	0.204041	1.000000	0.407577
9939	0.766617	0.754899	0.777646	0.808426	0.407577	1.000000

9.策略範例¶

9-1.均線多頭¶

In [ ]:

df=data.get('price:收盤價')
# 限定範圍
df=df[(df.index>'2015')&(df.index<'2019')]

cond1=df>df.rolling(5).mean()
cond2=df>df.rolling(10,min_periods=5).mean()
cond3=df>df.rolling(20,min_periods=10).mean()
cond4=df>df.rolling(60,min_periods=40).mean()

position=(df*(cond1&cond2&cond3&cond4))
position=position[position>0]
position=position.is_largest(10)

df=data.get('price:收盤價') # 限定範圍 df=df[(df.index>'2015')&(df.index<'2019')] cond1=df>df.rolling(5).mean() cond2=df>df.rolling(10,min_periods=5).mean() cond3=df>df.rolling(20,min_periods=10).mean() cond4=df>df.rolling(60,min_periods=40).mean() position=(df*(cond1&cond2&cond3&cond4)) position=position[position>0] position=position.is_largest(10)

Out[ ]:

	00632R	00635U	00638R	00726B	00732	00741B	00744B	0080	008201	1216	1225	1227	1232	1256	1259	1262	1264	1268	1301	1326	1338	1339	1464	1476	1477	1525	1527	1536	1537	1539	1541	1558	1565	1580	1589	1590	1597	1599	1702	1707	...	8114	8255	8279	8299	8341	8383	8406	8416	8422	8432	8433	8436	8437	8442	8444	8446	8450	8454	8462	8464	8466	8480	8489	8497	8913	8941	8942	9904	9910	9914	9917	9921	9924	9933	9937	9941	9943	9951	9958	9960
date
2015-01-05	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	...	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False
2015-01-06	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	...	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False
2015-01-07	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	...	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False
2015-01-08	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	...	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False
2015-01-09	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	...	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
2018-12-24	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	...	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False
2018-12-25	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	...	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False
2018-12-26	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	...	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False
2018-12-27	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	...	False	False	False	False	False	False	False	False	False	False	False	True	False	False	False	False	False	False	False	True	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False
2018-12-28	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	...	False	False	False	False	False	False	False	False	False	False	False	True	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False	False

981 rows × 314 columns

9-2.突破布林通道上緣¶

In [ ]:

df=data.get('price:收盤價')
mean_20=df.rolling(20).mean()
std_value=df.rolling(20).std()

up=mean_20+std_value*2

# 收盤價剛站上布林通道上緣
cond1=(df>up)&(df.shift()<up)
cond2=(df>5)&(df<200)
position=(cond1&cond2)
position=position[position>0]
position=position.is_smallest(20)

df=data.get('price:收盤價') mean_20=df.rolling(20).mean() std_value=df.rolling(20).std() up=mean_20+std_value*2 # 收盤價剛站上布林通道上緣 cond1=(df>up)&(df.shift()5)&(df<200) position=(cond1&cond2) position=position[position>0] position=position.is_smallest(20)

In [ ]: