def test_io(self):
df = stock.get_market_cap_by_ticker("20200101")
self.assertTrue(df.empty)
df = stock.get_market_cap_by_ticker("20200625")
self.assertFalse(df.empty)
for col in df.columns:
self.assertEqual(type(df[col].iloc[0]), np.int64)
def test_with_a_businessday(self):
df = stock.get_market_cap_by_ticker("20210104")
# 종가 시가총액 거래량 거래대금 상장주식수
# 티커
# 005930 83000 495491951650000 38655276 3185356823460 5969782550
# 000660 126000 91728297990000 7995016 994276505704 728002365
# 051910 889000 62756592927000 858451 747929748128 70592343
# 005935 74400 61222770480000 5455139 405685236800 822886700
# 207940 829000 54850785000000 182864 149889473000 66165000
self.assertIsInstance(df, pd.DataFrame)
self.assertEqual(len(df), 2531)
temp = np.isclose(df.iloc[0:5, 0], [83000, 126000, 889000, 74400, 829000])
self.assertEqual(temp.sum(), 5)
def get_latest_market_cap(today=None, market="ALL"):
if today == None:
today = datetime.datetime.today().strftime("%Y%m%d")
if DEBUG == True: print(get_latest_market_ohlcv.__name__, today)
today_datetime = datetime.datetime.strptime(today, "%Y%m%d")
for i in range(0, RETRY_COUNT):
latest_date = (today_datetime - datetime.timedelta(i)).strftime("%Y%m%d")
if DEBUG: print(latest_date)
result = stock.get_market_cap_by_ticker(latest_date, market)
if result.empty == False:
return result
return None
def __init__(self, collector_type):
# maria 연결 되어있는지 확인
self.is_maria = False
# dart api 연결 되어있는지 확인
self.is_dart = False
# chrome driver 연결 되어있는지 확인
self.is_chrome = False
# signet: 단기차입금, 장기차입금
# dart: 자본금, 무형자산
# fnguide: 나머지
# pykex: 상장주식수
self.collector_type = collector_type
# 기준 종목코드 2016년 상장되어있던 기업
stock_amount_2016 = stock.get_market_cap_by_ticker("20161229 ")[['상장주식수']]
self.code_list = list(stock_amount_2016.index)
def low_evaluation_list():
# 종목별 DIV/BPS/PER/EPS 조회
df = stock.get_market_fundamental_by_ticker("20200721")
df = df[(df['PER'] <= 9) & (df['PBR'] <= 1)]
df2 = stock.get_market_cap_by_ticker("20200721")
df3 = pd.merge(df, df2, left_index=True, right_index=True, how='outer')
df3 = df3[['종목명', 'DIV', 'BPS', 'PER', 'EPS', 'PBR', '시가총액', '거래량']]
df3['거래량'] = df3['거래량'].apply(lambda x: format(x, ','))
df3['시가총액'] = df3['시가총액'].apply(lambda x: x / 100000000).apply(int).apply(
lambda x: format(x, ',')) + ' 억'
df3['PBR'] = df3['PBR'].apply(lambda x: '%.2f' % x)
df3 = df3.sort_values(['EPS'], ascending=False)
print(df3.head(10))
def test_not_empty_result(self):
df = stock.get_market_ohlcv_by_date("20190225", "20190228", "000660")
self.assertTrue(df.empty)
df = stock.get_market_cap_by_date("20190101", "20190131", "005930")
self.assertTrue(df.empty)
df = stock.get_market_cap_by_ticker("20200625")
self.assertTrue(df.empty)
df = stock.get_market_price_change_by_ticker("20180301", "20180320")
self.assertTrue(df.empty)
df = stock.get_market_fundamental_by_ticker("20180305", "KOSPI")
self.assertTrue(df.empty)
df = stock.get_market_fundamental_by_date("20180301", "20180320", '005930')
self.assertTrue(df.empty)
def getMarketCapByTicker(self, date=None, market="ALL"):
if date is None:
return None
if not os.path.isdir(self.dirMarketCap):
os.makedirs(self.dirMarketCap)
filePath = self.dirMarketCap + '/' + \
'marketCap_%s_%s.bin' % (market, date)
if os.path.isfile(filePath):
with open(filePath, 'rb') as f:
marketCap = pickle.load(f)
return marketCap
rawMarketCap = None
try:
rawMarketCap = stock.get_market_cap_by_ticker(
date, market) # pandas form
except:
print("[KRX API update 가 필요합니다] stock.get_market_cap_by_ticker")
exit()
if rawMarketCap.empty == True:
print("[WARN] please check parameter `date`")
return None
resMarketCap = dict()
desc = "get marketCap"
for ticker in tqdm(rawMarketCap.index, desc):
tmpDict = dict()
corpName = stock.get_market_ticker_name(ticker)
tmpDict["종목명"] = corpName
for column in rawMarketCap.columns: # 종가, 시가총액, 거래량, 시가총액, 상장주식수
tmpDict[column] = str(
rawMarketCap.loc[ticker][column]).replace(" ", "")
resMarketCap[corpName] = tmpDict
# NOTE: 데이터 백업.
with open(filePath, 'wb') as f:
pickle.dump(resMarketCap, f)
return resMarketCap
def computeData(self, fnguide_tablename, signal_tablename, mh_tablename, jb_tablename, dg_tablename, res_tablename):
self._setMaria()
data = {}
std_date = '20201116'
today = '20210305'
res_tablename = res_tablename+std_date
years = ['2017', '2018', '2019', '2020']
for y in years:
data['fnguide{}'.format(y)] = self.maria.showData('select * from {}{}'.format(fnguide_tablename, y))
for y in years[:-1]:
data['new_signal{}'.format(y)] = self.maria.showData('select * from {}{}'.format(signal_tablename, y))
data['dart_mh'] = self.maria.showData('select * from {}'.format(mh_tablename))
data['dart_jb'] = self.maria.showData('select * from {}'.format(jb_tablename))
# 사업보고서가 발표 된다면 사용하지 않음
data['dart_dg_3q'] = self.maria.showData('select * from {}'.format(dg_tablename))
####################################
df = data['fnguide{}'.format(years[3])][['종목코드', '자산', '부채', '자본', '기말현금및현금성자산', '매출액', '영업이익발표기준', '당기순이익', 'GB']]
df = df.rename(columns={'자산': '자산총계{}'.format(years[3]), '부채': '부채총계{}'.format(years[3]), '자본': '자본총계{}'.format(years[3]), '기말현금및현금성자산': '현금성자산', '매출액': '매출액{}'.format(years[3]), '영업이익발표기준': '영업이익{}'.format(years[3]), '당기순이익': '당기순이익{}'.format(years[3])})
tmp = data['fnguide{}'.format(years[0])][['종목코드', '당기순이익', '영업이익발표기준', '자산']]
tmp = tmp.rename(columns={'당기순이익': '당기순이익{}'.format(years[0]), '영업이익발표기준': '영업이익{}'.format(years[0]),'자산': '자산총계{}'.format(years[0])})
df = pd.merge(df, tmp, on=['종목코드'])
tmp = data['fnguide{}'.format(years[1])][['종목코드', '당기순이익', '영업이익발표기준', '자본', '자산', '부채', '유동자산', '유동부채', '매출액']]
tmp = tmp.rename(columns={'당기순이익': '당기순이익{}'.format(years[1]), '영업이익발표기준': '영업이익{}'.format(years[1]), '자본': '자본총계{}'.format(years[1]), '자산': '자산총계{}'.format(years[1]), '부채': '부채총계{}'.format(years[1]), '유동자산': '유동자산{}'.format(years[1]), '유동부채': '유동부채{}'.format(years[1]), '매출액': '매출액{}'.format(years[1])})
df = pd.merge(df, tmp, on=['종목코드'])
tmp = data['fnguide{}'.format(years[2])][['종목코드', '당기순이익', '영업이익발표기준', '자본', '자산', '부채', '유동자산', '유동부채', '매출액']]
tmp = tmp.rename(columns={'당기순이익': '당기순이익{}'.format(years[2]), '영업이익발표기준': '영업이익{}'.format(years[2]), '자본': '자본총계{}'.format(years[2]), '자산': '자산총계{}'.format(years[2]), '부채': '부채총계{}'.format(years[2]), '유동자산': '유동자산{}'.format(years[2]), '유동부채': '유동부채{}'.format(years[2]), '매출액': '매출액{}'.format(years[2])})
df = pd.merge(df, tmp, on=['종목코드'])
# 사업보고서가 발표 된다면 해당년도로 사용
tmp = data['new_signal{}'.format(years[2])][['종목코드', '단기차입금', '장기차입금']]
df = pd.merge(df, tmp, on=['종목코드'])
tmp = data['dart_mh'][['종목코드', 'mh{}'.format(years[2])]]
tmp = tmp.rename(columns={'mh{}'.format(years[2]): '무형자산'})
df = pd.merge(df, tmp, on=['종목코드'])
tmp = data['dart_jb'][['종목코드', 'jb{}'.format(years[2])]]
tmp = tmp.rename(columns={'jb{}'.format(years[2]): '자본금'})
df = pd.merge(df, tmp, on=['종목코드'])
#############################################
df_An = df[df['GB'] == '12']
df_3q = df[df['GB'] == '09'].rename(columns={'당기순이익{}'.format(years[3]): '당기순이익{}_3Q'.format(years[3])})
tmp = data['dart_dg_3q'][['종목코드', 'dg{}_3Q'.format(years[1]), 'dg{}_3Q'.format(years[2])]]
tmp = tmp.rename(columns={'dg{}_3Q'.format(years[1]): '당기순이익{}_3Q'.format(years[1]), 'dg{}_3Q'.format(years[2]): '당기순이익{}_3Q'.format(years[2])})
df_3q = pd.merge(df_3q, tmp, on=['종목코드'])
tmp = df_3q[['종목코드', '당기순이익{}_3Q'.format(years[1]), '당기순이익{}_3Q'.format(years[2]), '당기순이익{}_3Q'.format(years[3])]]
tmp['rate{}'.format(years[2])] = (tmp['당기순이익{}_3Q'.format(years[2])] - tmp['당기순이익{}_3Q'.format(years[1])]) / tmp['당기순이익{}_3Q'.format(years[1])]
tmp['rate{}'.format(years[3])] = (tmp['당기순이익{}_3Q'.format(years[3])] - tmp['당기순이익{}_3Q'.format(years[2])]) / tmp['당기순이익{}_3Q'.format(years[2])]
tmp['당기순이익{}'.format(years[3])] = tmp['당기순이익{}_3Q'.format(years[2])] * (1 + (tmp['rate{}'.format(years[2])] + tmp['rate{}'.format(years[3])]) / 2)
df_3q = pd.merge(df_3q, tmp[['종목코드', '당기순이익{}'.format(years[3])]], on=['종목코드'])
df_3q = df_3q[list(df_An.columns)]
df = pd.concat([df_An, df_3q])
stock_amount_dates = ['20190329', '20200327', std_date]
for y in stock_amount_dates:
data['stock_amount_{}'.format(y)] = stock.get_market_cap_by_ticker(y)[['상장주식수', '시가총액']]
tmp = data['stock_amount_{}'.format(stock_amount_dates[0])].reset_index(drop=False)
tmp.columns = ['종목코드', '상장주식수{}'.format(years[1]), '시가총액{}'.format(years[1])]
tmp['상장주식수{}'.format(years[1])] = tmp['상장주식수{}'.format(years[1])] / 100000000
tmp['시가총액{}'.format(years[1])] = tmp['시가총액{}'.format(years[1])] / 100000000
df = pd.merge(df, tmp, on=['종목코드'])
tmp = data['stock_amount_{}'.format(stock_amount_dates[1])].reset_index(drop=False)
tmp.columns = ['종목코드', '상장주식수{}'.format(years[2]), '시가총액{}'.format(years[2])]
tmp['상장주식수{}'.format(years[2])] = tmp['상장주식수{}'.format(years[2])] / 100000000
tmp['시가총액{}'.format(years[2])] = tmp['시가총액{}'.format(years[2])] / 100000000
df = pd.merge(df, tmp, on=['종목코드'])
tmp = data['stock_amount_{}'.format(stock_amount_dates[2])].reset_index(drop=False)
tmp.columns = ['종목코드', '상장주식수{}'.format(years[3]), '시가총액{}'.format(years[3])]
tmp['상장주식수{}'.format(years[3])] = tmp['상장주식수{}'.format(years[3])] / 100000000
tmp['시가총액{}'.format(years[3])] = tmp['시가총액{}'.format(years[3])] / 100000000
#df = pd.merge(df, tmp, on=['종목코드'])
df = pd.merge(df, tmp[['종목코드', '시가총액{}'.format(years[3])]], on=['종목코드'])
tmp = self.maria.showData('select * from stocks_202103')
tmp['상장주식수{}'.format(years[3])] = (tmp['보통주'] + tmp['우선주']) / 100000000
df = pd.merge(df, tmp[['종목코드', '상장주식수{}'.format(years[3])]], on=['종목코드'])
for y in years[1:]:
df['EPS{}'.format(y)] = df['당기순이익{}'.format(y)] / df['상장주식수{}'.format(y)]
df1 = pd.DataFrame(stock.get_market_ohlcv_by_ticker(std_date)['종가'])
df1['Size'] = [0] * df1.shape[0]
df2 = pd.DataFrame(stock.get_market_ohlcv_by_ticker(std_date, market='KOSDAQ')['종가'])
df2['Size'] = [1] * df2.shape[0]
df3 = pd.DataFrame(stock.get_market_ohlcv_by_ticker(std_date, market='KONEX')['종가'])
df3['Size'] = [2] * df3.shape[0]
price = pd.concat([df1, df2, df3])
price = price.reset_index(drop=False).rename(columns={'티커': '종목코드', '종가': '주식가격'})
df = pd.merge(df, price, on='종목코드')
df['순자산가치'] = (df['자본총계{}'.format(years[3])] - df['무형자산']) / df['상장주식수{}'.format(years[3])]
df['순손익가치'] = ((df['EPS{}'.format(years[1])] + df['EPS{}'.format(years[2])]*2 + df['EPS{}'.format(years[3])]*3)/6)*10
df['본질가치'] = (df['순자산가치'] + (df['순손익가치']*1.5))/2.5
df['보충적가치'] = (df['순손익가치']*2 + df['순자산가치']*3) / 5
df['주당기업가치'] = (df['상장주식수{}'.format(years[3])]*df['주식가격']+df['단기차입금']+df['장기차입금']-df['현금성자산'])/df['상장주식수{}'.format(years[3])]
df['주당가치분석'] = (df['본질가치'] + df['주당기업가치']) / 2
cols = list(df.columns)
cols.remove('GB')
df = df[cols]
print(df.columns)
print(len(list(df.columns)))
self.maria.createTable(res_tablename, list(df.columns), ['VARCHAR(20)'] + ['FLOAT']*len(cols))
for i in range(df.shape[0]):
try:
self.maria.insertData(res_tablename, tuple(list(df.iloc[i])))
self.maria.commitDB()
except:
continue
def fnguideCollect(self):
self._setMaria()
cols = ['날짜', '자산', '부채', '자본', '기초현금및현금성자산', '기말현금및현금성자산',
'유동자산계산에 참여한 계정 펼치기', '유동부채계산에 참여한 계정 펼치기',
'매출액', '영업이익', '영업이익(발표기준)', '당기순이익']
changed_cols = [c.replace('계산에 참여한 계정 펼치기', '') for c in cols]
changed_cols = [c.replace('(', '') for c in changed_cols]
changed_cols = [c.replace(')', '') for c in changed_cols]
fn_tablename = 'fnguide'
years = ['2017', '2018', '2019', '2020']
for batch in range(20, 2500, 10):
data = {}
data['2017'] = []
data['2018'] = []
data['2019'] = []
data['2020'] = []
for code in self.code_list[batch:batch + 10]:
df = self.getFN(code)
if df is None:
self.writeError('error_fnguide', code, 'No page')
continue
try:
fs = df.columns[0] # 연결 or 개별?
df = df.rename(columns={df.columns[0]: "날짜"})[cols]
df['FS'] = [fs] * df.shape[0]
df['GB'] = [date[-2:] for date in list(df['날짜'])]
df.columns = changed_cols + ['FS', 'GB']
for i, y in enumerate(years):
data[y].append([code] + list(df.iloc[i]))
except:
self.writeError('error_fnguide', code, 'No cols')
continue
stock_amount_dates = ['20171228', '20181228', '20191230', '20201230', '20200929']
for i, y in enumerate(years[:-1]):
data[y] = pd.DataFrame(data[y])
data[y].columns = ['티커'] + changed_cols + ['FS', 'GB']
tmp = data[y][data[y]['GB'] == '12']
tmp = tmp.set_index('티커')[changed_cols[1:] + ['FS', 'GB']]
data['stock_amount_{}'.format(y)] = stock.get_market_cap_by_ticker(stock_amount_dates[i])[['상장주식수']]
tmp = pd.merge(tmp, data['stock_amount_{}'.format(y)], left_index=True, right_index=True)[changed_cols[1:] + ['상장주식수', 'FS', 'GB']].fillna(-9999999)
for j in range(tmp.shape[0]):
self.maria.insertData('{}{}'.format(fn_tablename, y), tuple(
[tmp.index[j]] + list(map(int, list(tmp.iloc[j][tmp.columns[:-2]]))) + [
tmp['FS'].iloc[j][-3:-1]] + [tmp['GB'].iloc[j]]))
self.maria.commitDB()
i += 1
y = years[i]
data[y] = pd.DataFrame(data[y])
data[y].columns = ['티커'] + changed_cols + ['FS', 'GB']
tmp = data[y][data[y]['GB'] == '12']
tmp = tmp.set_index('티커')[changed_cols[1:] + ['FS', 'GB']]
data['stock_amount_{}'.format(y)] = stock.get_market_cap_by_ticker(stock_amount_dates[i])[['상장주식수']]
tmp = pd.merge(tmp, data['stock_amount_{}'.format(y)], left_index=True, right_index=True)[changed_cols[1:] + ['상장주식수', 'FS', 'GB']].fillna(-9999999)
for j in range(tmp.shape[0]):
self.maria.insertData('{}{}'.format(fn_tablename, y), tuple(
[tmp.index[j]] + list(map(int, list(tmp.iloc[j][tmp.columns[:-2]]))) + [
tmp['FS'].iloc[j][-3:-1]] + [tmp['GB'].iloc[j]]))
self.maria.commitDB()
tmp = data[y][data[y]['GB'] == '09']
tmp = tmp.set_index('티커')[changed_cols[1:] + ['FS', 'GB']]
data['stock_amount_{}'.format(y)] = stock.get_market_cap_by_ticker(stock_amount_dates[i+1])[['상장주식수']]
tmp = pd.merge(tmp, data['stock_amount_{}'.format(y)], left_index=True, right_index=True)[changed_cols[1:] + ['상장주식수', 'FS', 'GB']].fillna(-9999999)
for j in range(tmp.shape[0]):
self.maria.insertData('{}{}'.format(fn_tablename, y), tuple(
[tmp.index[j]] + list(map(int, list(tmp.iloc[j][tmp.columns[:-2]]))) + [
tmp['FS'].iloc[j][-3:-1]] + [tmp['GB'].iloc[j]]))
self.maria.commitDB()
from pykrx import stock
import pandas as pd
tickers = stock.get_market_ticker_list()
# 종목별 DIV/BPS/PER/EPS 조회
df = stock.get_market_fundamental_by_ticker("20200721")
df = df[(df['PER'] <= 9) & (df['PBR'] <= 1)]
df2 = stock.get_market_cap_by_ticker("20200721")
df3 = pd.merge(df, df2, left_index=True, right_index=True, how='outer')
#df3= df3.sort_values(['EPS'], ascending=False)
#df.to_csv('df.csv', encoding='utf-8')
#df2.to_csv('df2.csv', encoding='utf-8')
#df3.sort_index().to_csv('df3.csv', encoding='utf-8')
print(df.sort_index())
print(df2.sort_index())
df3 = df3[['종목명', 'DIV', 'BPS', 'PER', 'EPS', 'PBR', '시가총액', '거래량']]
for c in df3[['시가총액', '거래량']]:
df3[c] = df3[c].apply(lambda x: format(x, ','))
df3 = df3.sort_values(['EPS'], ascending=False)
print(df3.head(10))
'''
market_sum = []
quant = []
for ticker in df.loc[:10, '티커']:
ticker_info = stock.get_market_cap_by_date("20200721", "20200721", ticker)
market_sum.append(ticker_info.iloc[0, 0])
def test_with_a_holiday(self):
df_0 = stock.get_market_cap_by_ticker("20210103")
df_1 = stock.get_market_cap_by_ticker("20201230")
# 해당 연휴에 상폐된 종목이 없기 때문에 두 index를 비교해 봄
result = set(df_0.index) - set(df_1.index)
self.assertEqual(len(result), 0)
def test_with_a_holiday(self):
df_0 = stock.get_market_cap_by_ticker("20210103")
df_1 = stock.get_market_cap_by_ticker("20201230")
same = (df_0 == df_1).all(axis=None)
self.assertTrue(same)
def kospi_n_codes(at: date, n: int) -> List[str]:
# noinspection PyTypeChecker
return [
cap[0] for cap in stock.get_market_cap_by_ticker(
date=__date_to_str(at)).iterrows()
][:n]
def getPykrxData():
stock_amount = stock.get_market_cap_by_ticker("20200929")[['상장주식수']]
return stock_amount
