def All_stock():
df = stock.get_market_ohlcv_by_ticker('20210217', market='KOSPI')
list_of_index = [item for item in df.index]
ticker = sorted(list_of_index) #종목코드 오름차순
ticker_name = [stock.get_market_ticker_name(i) for i in ticker] # 종목명 오름차순
return ticker, ticker_name
def krx_price_query(request):
"""
가격 정보.
:param request:
:return:
"""
render_dict = get_render_dict("krx_price_query")
if request.POST:
query = request.POST["query"]
render_dict["query"] = query
realtime_result = get_xml_request(krx_price_query_url.format(query))
if realtime_result:
krx_realtime(realtime_result, query, render_dict)
statement_result = get_xml_request(
krx_statement_query_url.format(query))
if statement_result:
krx_statement(statement_result, render_dict)
# df = stock.get_market_ohlcv_by_date("20200426", "20200426", query)
# render_dict["market"] = df
ticker_name = stock.get_market_ticker_name(query)
render_dict["ticker_name"] = ticker_name
stocks = models.Stock.objects.all().order_by("code")
render_dict["stocks"] = stocks
return render(request, "book/investment/krx_price_query.html", render_dict)
def pykrx_scratch(date_Start, date_End):
print("Reading Daily Chart ... {} - {}".format(date_Start, date_End))
# create main folder
if not os.path.exists(Krx_Char_folder_path):
os.mkdir(Krx_Char_folder_path)
# ticker scratch
for ticker in stock.get_market_ticker_list(market="ALL"):
stock_name = stock.get_market_ticker_name(ticker)
stock_folder_name = stock_name + '_' + ticker
#print(stock_name, ticker)
df = stock.get_market_ohlcv_by_date(date_Start, date_End, ticker)
df = df.reset_index()
#print(len(df))
df.insert(5, '종가2', df['종가'])
# folder check
if not os.path.exists(Krx_Char_folder_path + '/' + stock_name):
os.mkdir(Krx_Char_folder_path + '/' + stock_name)
df.to_csv(Krx_Char_folder_path + '/' + stock_name + '/' + ticker +
'.csv',
sep=',',
na_rep='0',
index=False,
header=False)
print('{} Daily chart is written! ==== ticker is : {}'.format(
stock_name, ticker))
print('Scratching daily chart is done!')
def getTickerNameALL(self):
logger.info("GetStockList - getTickerNameALL")
for stockTicker in self.ALL:
try:
self.tickerToName[stockTicker] = stock.get_market_ticker_name(stockTicker)
except:
self.tickerToName[stockTicker] = stockTicker
logger.critical(f"GetStockList - getTickerNameALL; No name retrieved, ticker : {stockTicker}")
def save_market_code_name(self, date, market):
item_codes = stock.get_market_ticker_list(date, market=market)
for item_code in item_codes:
item_name = stock.get_market_ticker_name(item_code)
item_code = int(item_code)
data = {
"stock_item_name": item_name,
"stock_item_code": item_code,
"stock_market_name": market
}
req = CreateItemListReq()
req.set_param(data)
def post_item_list(self, tickers_in_market: dict):
req = CreateItemListReq()
for market_name in tickers_in_market.keys():
tickers = tickers_in_market[market_name]
for ticker in tickers:
item_name = stock.get_market_ticker_name(ticker)
params = {'stock_item_name': item_name,
'stock_item_code': ticker,
'stock_market_name': market_name}
req.set_param(params)
res = req.send_post()
print(res.json())
def getCorpCodeByCorpName(self, corpName):
if self.useCorpCodeXml == True:
for corpList in self.corpList:
if corpList.findtext("corp_name") == corpName:
return corpList.findtext("corp_code")
else:
for ticker in self.tickerList:
name = stock.get_market_ticker_name(ticker)
if name == corpName:
return ticker
return None
def __collect_tickers(market='KOSPI'):
"""
stock.get_market_ticker_list()
>> ['095570', '006840', '027410', '282330', '138930', ...]
stock.get_market_ticker_name(ticker)
>> SK하이닉스
"""
tickers = dict()
today = datetime.today().strftime("%Y%m%d")
for ticker in stock.get_market_ticker_list(today, market):
ticker_name = stock.get_market_ticker_name(ticker)
tickers[ticker_name] = ticker
return tickers
def find_all_codes(fromdate: date, todate: date):
tempd = fromdate
codes = {}
while True:
codes.update({code: None for code in pykrx_stock.get_market_ticker_list(_date_to_str(tempd), market='ALL')})
if tempd > todate:
break
tempd += timedelta(days=365)
for code in codes:
name = pykrx_stock.get_market_ticker_name(code)
if isinstance(name, str):
codes.update({code: name})
return codes
def getMarketOhlcvByTicker(self, date=None, market="ALL"):
if date == None:
return None
# NOTE: 백업 데이터 존재 여부 확인.
if not os.path.isdir(self.dirMarketOhlcv):
os.makedirs(self.dirMarketOhlcv)
filePath = self.dirMarketOhlcv + \
'/marketOhlcv_%s_%s.bin' % (market, date)
if os.path.isfile(filePath):
# NOTE: file 이 있고, corpName 에 해당하는 데이터 있는지 확인.
with open(filePath, 'rb') as f:
marketOhlcv = pickle.load(f)
return marketOhlcv
rawOhlcvList = None
try:
rawOhlcvList = stock.get_market_ohlcv_by_ticker(date, market)
except:
print("[param] date 를 확인해주세요.")
print(
"[위 사항이 아니라면] [KRX API update 가 필요합니다] stock.get_market_ohlcv_by_ticker")
exit()
# NOTE: rawOhlcvList 데이터는 pandas.DataFrame 인데, DaraFrame 전체가 empty 면 rawOhlcvList.empty 는 True 를 리턴 함.
# [refer] https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.empty.html
if rawOhlcvList.empty == True:
return None
resOhlcv = dict()
desc = "get ohlcv"
for ticker in tqdm(rawOhlcvList.index, desc):
corpName = stock.get_market_ticker_name(ticker)
tmpDict = dict()
for column in rawOhlcvList.columns:
tmpDict[column] = str(rawOhlcvList.loc[ticker][column])
resOhlcv[corpName] = tmpDict
# NOTE: 데이터 백업.
with open(filePath, 'wb') as f:
pickle.dump(resOhlcv, f)
return resOhlcv
def getMarketFundamentalByTicker(self, date=None, market="ALL"):
if date == None:
return None
# NOTE: 백업 데이터 존재 여부 확인.
if not os.path.isdir(self.dirMarketFundamental):
os.makedirs(self.dirMarketFundamental)
filePath = self.dirMarketFundamental + \
'/marketFundamental_%s_%s.bin' % (market, date)
if os.path.isfile(filePath):
# NOTE: file 이 있고, corpName 에 해당하는 데이터 있는지 확인.
with open(filePath, 'rb') as f:
marketFundamental = pickle.load(f)
return marketFundamental
rawFundamentalList = None
try:
rawFundamentalList = stock.get_market_fundamental_by_ticker(
date, market) # pandas form
except:
print("[KRX API update 가 필요합니다] stock.get_market_fundamental_by_ticker")
exit()
if rawFundamentalList.empty == True:
print("[WARN] please check parameter `date`")
return None
resFundamental = dict()
desc = "get fundamental"
for ticker in tqdm(rawFundamentalList.index, desc):
corpName = stock.get_market_ticker_name(ticker)
tmpDict = dict()
for column in rawFundamentalList.columns:
tmpDict[column] = str(
rawFundamentalList.loc[ticker][column]).replace(" ", "")
resFundamental[corpName] = tmpDict
# NOTE: 데이터 백업.
with open(filePath, 'wb') as f:
pickle.dump(resFundamental, f)
return resFundamental
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
class krxModule():
tickerList = stock.get_market_ticker_list()
nameList = map(lambda x: stock.get_market_ticker_name(x), tickerList)
nameMap = dict(zip(nameList, tickerList))
#key==index
#value==dictionary{colums:value}
def to_dict(self, df):
return {
df.iloc[i].name: df.iloc[i, :].to_dict()
for i in range(len(df))
}
def chartJSON(self, dic):
tmp_dic = {ticker["종목명"]: ticker["등락률"] for _, ticker in dic.items()}
labels = list(tmp_dic.keys())
datas = list(tmp_dic.values())
rand = [[
np.random.randint(0, 255),
np.random.randint(0, 255),
np.random.randint(0, 255)
] for _ in range(len(labels))]
backgroundColors = [
'rgba(%d,%d,%d,0.2)' % (rand[i][0], rand[i][1], rand[i][2])
for i in range(len(labels))
]
borderColors = [
'rgba(%d,%d,%d,1)' % (rand[i][0], rand[i][1], rand[i][2])
for i in range(len(labels))
]
json_ = {
'type': 'bar',
'data': {
'labels':
labels,
'datasets': [{
'label': '등락률',
'data': datas,
'backgroundColor': backgroundColors,
'borderColor': borderColors,
'borderWidth': '1'
}]
},
'options': {
'responsive': False,
'scales': {
'yAxes': [{
'ticks': {
'beginAtZero': True
}
}]
},
}
}
return json_
def getName_by_ticker(self, ticker):
return self.nameMap[ticker]
def getTop10(self, start, end):
df = stock.get_market_price_change_by_ticker(start, end)
return self.to_dict(df.sort_values("등락률", ascending=False).head(10))
def news_finder(stock_id, start_date):
start_date = start_date
stock_name = stock.get_market_ticker_name(stock_id)
# 네이버 금융 뉴스 섹션에 접속
driver = webdriver.Chrome(executable_path="./chromedriver.exe")
url = "https://finance.naver.com/news/"
driver.get(url)
driver.find_element_by_xpath(
'//*[@id="newsMainTop"]/div/div[2]/form/div/input').click(
) # 검색창 초기화
driver.find_element_by_xpath(
'//*[@id="newsMainTop"]/div/div[2]/form/div/input').send_keys(
stock_name) # 주식 이름 검색창에 입력
driver.find_element_by_xpath(
'//*[@id="newsMainTop"]/div/div[2]/form/div/a').click(
) # 검색 아이콘 클릭해서 실행
now = datetime.datetime.now()
today_date = str(now.strftime('%Y-%m-%d'))
result_url = driver.current_url
new_url = result_url + "&sm=title.basic&pd=4&stDateStart=" + start_date + "&stDateEnd=" + today_date # 제목에서만 설정 & Start 날짜 직접 설정 가능!
driver.get(new_url)
# 몇 페이지까지 존재하는지 알아보자.
bs_obj = bs4.BeautifulSoup(driver.page_source,
"html.parser") # 뷰티풀숩 object 생성
last = bs_obj.find("td", {"class": "pgRR"})
a = last.find('a', href=True)
last_page_num = int(a['href'].split('page=')[1])
# last_page_num 개의 페이지 URL을 pages라는 리스트에 담아보자.
page_numbering = list(range(1, last_page_num + 1))
pages = []
for i in range(1, last_page_num):
pages.append(new_url + "&page=" + str(i))
# 빈 데이터프레임 생성
df = pd.DataFrame(columns=("date", "title", "content"))
# 빈 데이터프레임에 크롤링한 기사 본문 내용들을 채워넣기
for page in pages:
driver.get(page)
bs_obj = bs4.BeautifulSoup(driver.page_source, "html.parser")
# 깔끔한 작업을 위해 필요한 영역만 남기고 나머지는 무시하자
news_list = bs_obj.find("dl", {"class": "newsList"})
news_titles_1 = news_list.find_all("dt",
{"class": "articleSubject"})
news_titles_2 = news_list.find_all(
"dd", {"class": "articleSubject"
}) # 네이버 뉴스는 썸네일이 있는 뉴스와 없는 뉴스의 태그가 다르게 설정되어 있네요.
news_titles = news_titles_1 + news_titles_2 # 그래서 작업을 2번 한 후에 합치는 방식을 썼습니다.
del news_titles_1, news_titles_2 # 필요없는 변수 제거
for title in news_titles:
temp = title.find('a', href=True)
news_url = "https://finance.naver.com" + temp['href']
driver.get(news_url) # 개별 뉴스들을 클릭하는 동작을 실행시킨다.
bs_obj = bs4.BeautifulSoup(driver.page_source, "html.parser")
if len(driver.window_handles) > 1: # 팝업창이 있는 경우
time.sleep(1)
driver.switch_to_window(driver.window_handles[1])
driver.close() # 팝업창 종료
driver.switch_to_window(
driver.window_handles[0]) # 원래창으로 복귀
else: # 팝업창이 없는 경우
# 기사 제목 크롤링
title = driver.find_element_by_xpath(
'//*[@id="contentarea_left"]/div[2]/div[1]/div[2]/h3'
).text
# 기사 업로드 날짜 크롤링
date = driver.find_element_by_xpath(
'//*[@id="contentarea_left"]/div[2]/div[1]/div[2]/div/span'
).text
# 기사 본문 크롤링
content = collect_data.nlp_news(
driver.find_element_by_xpath(
'//*[@id="content"]').text)
df.loc[len(df)] = [date, title, content] # 빈 데이터프레임에 행 추가
# id와 name이라는 column을 만들고 해당 값으로 동일하게 채워넣는다.
df['id'] = stock_id
df['name'] = stock_name
driver.close()
return df
def test_io_in_kosdaq_market(self):
tickers = stock.get_market_ticker_list("20190225", market="KOSDAQ")
for ticker in tickers:
name = stock.get_market_ticker_name(ticker)
self.assertIsInstance(name, str)
self.assertNotEqual(len(name), 0)
def test_io(self):
name = stock.get_market_ticker_name("000660")
self.assertIsInstance(name, str)
self.assertNotEqual(len(name), 0)
simulation_size = 10
num_layers = 2
size_layer = 256
timestamp = 10
epoch = 500
dropout_rate = 0.8
test_size = 30
learning_rate = 0.0001
date_Start = '20200101'
stock_input = "코리아센터"
date_End = datetime.today().strftime("%Y%m%d")
print("Daily candle dates {} - {}".format(date_Start, date_End))
for ticker in stock.get_market_ticker_list(market="ALL"):
stock_name = stock.get_market_ticker_name(ticker)
if stock_name == stock_input:
print('Found : {}, ticker : {}'.format(stock_input, ticker))
korea_center_ticker = ticker
korea_center_name = stock_input
df_korea = stock.get_market_ohlcv_by_date(date_Start, date_End,
korea_center_ticker)
df_korea = df_korea.reset_index()
print(len(df_korea))
df_korea.insert(5, '종가2', df_korea['종가'])
#df = pd.read_csv('../dataset/GOOG-year.csv')
#df_korea.rename(columns=df.columns)
#print(df_korea.head())
#print(df.head())
print(df_korea.iloc[:, 4:5].tail())
def get_name(code: str):
if code not in name_cache:
name_cache.update({code: stock.get_market_ticker_name(code)})
return name_cache.get(code)
from pykrx import stock
import json
# 회사 이름에 대응되는 ticker라는 것이 있는데 그걸 받아온다
tickers = stock.get_market_ticker_list()
# print(tickers)
myJSON = {}
for ticker in tickers:
myJSON[ticker] = stock.get_market_ticker_name(ticker)
# name = stock.get_market_ticker_name("000540")
with open("tickerlist.json", "w") as JSON:
json.dump(myJSON, JSON, ensure_ascii=False)
# with open("tickerlist.json", "r") as JSON:
# myJSON = json.load(JSON)
print(myJSON)
ohlcv = stock.get_market_ohlcv_by_date("20200501", "20200510", "000660")
print(ohlcv)
df = stock.get_market_price_change_by_ticker("20200501", "20200510")
print(df)
