# ** 5 Python for Finance** 파이썬을 활용한 금융분석
1 금융 시계열 데이터 결측치 처리 / 보간법
https://github.com/pydata/pandas-datareader
- pandas를 활용한 데이터 정렬 및 보간법
- 금융데이터 크롤링 모듈 설치방법 (2018.06 기준)
- ! pip install git+https://github.com/pydata/pandas-datareader.git --upgrade
In [1]:
# from pandas_datareader import get_data_yahoo
# start_date = '2013-01-01'
# codes = ['KO', 'MS', '005930.KS', '000660.KS']
# result = {}
# for code in codes:
# result[code] = get_data_yahoo(code, start_date).Close
# import pandas as pd
# prices = pd.DataFrame(result)
# prices = prices.fillna(method='ffill')
# prices = prices.dropna()
# prices.to_csv('./data/stocks.csv', encoding="ms949")
# prices.tail(3)
In [2]:
# 저장된 CSV 파일을 사용한다
import pandas as pd
prices = pd.read_csv('./data/stocks.csv')
prices = prices.set_index('Date')
prices.head(3)
prices.index = pd.DatetimeIndex(prices.index)
prices.info()
<class 'pandas.core.frame.DataFrame'> DatetimeIndex: 1410 entries, 2013-01-02 to 2018-06-11 Data columns (total 4 columns): KO 1410 non-null float64 MS 1410 non-null float64 005930.KS 1410 non-null float64 000660.KS 1410 non-null float64 dtypes: float64(4) memory usage: 55.1 KB
### **01 가격 차이가 큰 시계열 데이터 비교하기** 데이터 정규화 / 비정규화
In [3]:
%matplotlib inline
prices.plot()
Out[3]:
<matplotlib.axes._subplots.AxesSubplot at 0x7fb42f67cfd0>
In [4]:
price = prices['005930.KS']
price[-5:]
Out[4]:
Date 2018-06-05 51300.0 2018-06-06 51300.0 2018-06-07 50600.0 2018-06-08 49650.0 2018-06-11 49900.0 Name: 005930.KS, dtype: float64
In [5]:
price.index
Out[5]:
DatetimeIndex(['2013-01-02', '2013-01-03', '2013-01-04', '2013-01-07',
'2013-01-08', '2013-01-09', '2013-01-10', '2013-01-11',
'2013-01-14', '2013-01-15',
...
'2018-05-29', '2018-05-30', '2018-05-31', '2018-06-01',
'2018-06-04', '2018-06-05', '2018-06-06', '2018-06-07',
'2018-06-08', '2018-06-11'],
dtype='datetime64[ns]', name='Date', length=1410, freq=None)
In [6]:
price.index = pd.DatetimeIndex(price.index)
price.index
Out[6]:
DatetimeIndex(['2013-01-02', '2013-01-03', '2013-01-04', '2013-01-07',
'2013-01-08', '2013-01-09', '2013-01-10', '2013-01-11',
'2013-01-14', '2013-01-15',
...
'2018-05-29', '2018-05-30', '2018-05-31', '2018-06-01',
'2018-06-04', '2018-06-05', '2018-06-06', '2018-06-07',
'2018-06-08', '2018-06-11'],
dtype='datetime64[ns]', name='Date', length=1410, freq=None)
In [7]:
price["2013-10-03"] / price["2013-08-01"] -1
Out[7]:
0.10781250000000009
In [8]:
returns = price.pct_change()
ret_price = (1 + returns).cumprod()
ret_price[0] = 1
%matplotlib inline
ret_price.plot(figsize=(12,3))
Out[8]:
<matplotlib.axes._subplots.AxesSubplot at 0x7fb42d028320>
In [9]:
for col in prices.columns:
pct_returns = prices[col].pct_change()
pct_returns = (1 + pct_returns).cumprod()
pct_returns[0] = 1
pct_returns.plot(legend=True)
### **02 누적 수익률 계산 ** .pct_change().cumprod()
In [10]:
month_returns = price.resample('BM').last().pct_change()
month_returns['2016']
Out[10]:
Date 2016-01-29 -0.087302 2016-02-29 0.024348 2016-03-31 0.113752 2016-04-29 -0.051067 2016-05-31 0.037751 2016-06-30 0.102941 2016-07-29 0.080000 2016-08-31 0.052632 2016-09-30 -0.013580 2016-10-31 0.025657 2016-11-30 0.065284 2016-12-30 0.032073 Freq: BM, Name: 005930.KS, dtype: float64
In [11]:
m_ret = (1 + month_returns).cumprod()
m_ret['2016']
Out[11]:
Date 2016-01-29 0.794199 2016-02-29 0.813536 2016-03-31 0.906077 2016-04-29 0.859807 2016-05-31 0.892265 2016-06-30 0.984116 2016-07-29 1.062845 2016-08-31 1.118785 2016-09-30 1.103591 2016-10-31 1.131906 2016-11-30 1.205801 2016-12-30 1.244475 Freq: BM, Name: 005930.KS, dtype: float64
In [12]:
import matplotlib.pyplot as plt
m_ret.plot(figsize=(12,3))
plt.axhline(y = 1, linestyle = '--', color = 'k')
plt.show()
### **03 중간 배당데이터 추가하기** 재귀적 연산
In [13]:
month_returns['2016-03-31'] += 0.003
In [14]:
for i in range(4):
i += 10000
print(i)
10000 10001 10002 10003
### **01 Introduction**
In [15]:
prices.head(3)
Out[15]:
| KO | MS | 005930.KS | 000660.KS | |
|---|---|---|---|---|
| Date | ||||
| 2013-01-02 | 37.599998 | 19.620001 | 31520.0 | 26600.0 |
| 2013-01-03 | 37.599998 | 19.580000 | 30860.0 | 26650.0 |
| 2013-01-04 | 37.660000 | 20.190001 | 30500.0 | 26350.0 |
In [16]:
prices.describe()
Out[16]:
| KO | MS | 005930.KS | 000660.KS | |
|---|---|---|---|---|
| count | 1410.000000 | 1410.000000 | 1410.000000 | 1410.000000 |
| mean | 42.100099 | 35.681879 | 32640.000000 | 45578.581560 |
| std | 2.407227 | 9.298308 | 9609.622299 | 17468.595733 |
| min | 36.840000 | 19.580000 | 21340.000000 | 23700.000000 |
| 25% | 40.509998 | 28.882501 | 25840.000000 | 31800.000000 |
| 50% | 41.990002 | 33.555000 | 28560.000000 | 41850.000000 |
| 75% | 43.787500 | 42.627501 | 38205.000000 | 49787.500000 |
| max | 48.529999 | 58.910000 | 57220.000000 | 95300.000000 |
In [17]:
prices.describe().loc[ ['mean', 'std'] ,:]
Out[17]:
| KO | MS | 005930.KS | 000660.KS | |
|---|---|---|---|---|
| mean | 42.100099 | 35.681879 | 32640.000000 | 45578.581560 |
| std | 2.407227 | 9.298308 | 9609.622299 | 17468.595733 |
In [18]:
prices.agg(['mean', 'std'])
Out[18]:
| KO | MS | 005930.KS | 000660.KS | |
|---|---|---|---|---|
| mean | 42.100099 | 35.681879 | 32640.000000 | 45578.581560 |
| std | 2.407227 | 9.298308 | 9609.622299 | 17468.595733 |
### **02 주가간의 상관계수 분석**
In [19]:
prices_corr = prices.pct_change().dropna().corr()
prices_corr
Out[19]:
| KO | MS | 005930.KS | 000660.KS | |
|---|---|---|---|---|
| KO | 1.000000 | 0.277816 | 0.041527 | 0.037768 |
| MS | 0.277816 | 1.000000 | 0.074325 | 0.099111 |
| 005930.KS | 0.041527 | 0.074325 | 1.000000 | 0.343296 |
| 000660.KS | 0.037768 | 0.099111 | 0.343296 | 1.000000 |
In [20]:
%matplotlib inline
import seaborn as sns
sns.heatmap(prices_corr)
Out[20]:
<matplotlib.axes._subplots.AxesSubplot at 0x7fb421ea12b0>
