The Relationship between news and stocks 3

회의록

- CV 와 NLP 주제 고민

- CV로 주제 선정시 Style GAN의 필요 성능이 너무 높아 불가능 할거라고 판단

- NLP 주제인 뉴스기사로 인한 주가 등락 예측 주제로 선정

- 네이버 주식 사이트에서 KOSPI 200 기업의 주식시세와 관련 뉴스 데이터 수집

코드

클래스 생성

class Crolling:
	def __init__(self):
		self.conn = pymysql.connect(user    = 'stocks',
									passwd  = '',
									host    = "",
									port    =  3306,
									db      = 'Data',
									charset = 'utf8')

		self.cur = self.conn.cursor()

		self.driver = webdriver.Chrome(ChromeDriverManager().install())
		self.driver.implicitly_wait(3)
		self.driver.maximize_window()

KOSPI 200 기업 고유 아이디 수집

	def Stock_ID(self):
		for i in range(1, 21):
			url = f'https://finance.naver.com/sise/entryJongmok.naver?&page={i}'
			self.driver.get(url)
			self.driver.implicitly_wait(5)
			html = self.driver.page_source 
			soup = BeautifulSoup(html, 'html.parser')
			for j in range(3, 13):
				id = [i['href'].split("=")[1] for i in soup.select(f'body > div > table.type_1 > tbody > tr:nth-child({j}) > td.ctg > a')]
				name = [i.text for i in soup.select(f'body > div > table.type_1 > tbody > tr:nth-child({j}) > td.ctg > a')]
				sql = (id, name)
				self.cur.execute('INSERT IGNORE INTO Stock_ID (id, name) VALUES (%s ,%s)', sql)
				self.conn.commit()
			print(f'{i}/20 페이지 완료')
		self.driver.quit()

주식 시세 수집 ( 3월 31 ~ 7월 18일 )

	def Stock_Price(self):
		self.cur.execute('SELECT id FROM Stock_ID;')
		stock_id = self.cur.fetchall()
		for id in stock_id:
			for i in range(1, 21):
				url = f'https://finance.naver.com/item/sise_day.naver?code={id[0]}&page={i}'
				self.driver.get(url)
				self.driver.implicitly_wait(5)
				html = self.driver.page_source 
				soup = BeautifulSoup(html, 'html.parser')
				for j in [3,4,5,6,7,11,12,13,14,15]:
					date = soup.select_one(f'body > table.type2 > tbody > tr:nth-child({j}) > td:nth-child(1) > span').text
					closing_price = soup.select_one(f'body > table.type2 > tbody > tr:nth-child({j}) > td:nth-child(2) > span').text.replace(',','')
					market_price = soup.select_one(f'body > table.type2 > tbody > tr:nth-child({j}) > td:nth-child(4) > span').text.replace(',','')
					high_price = soup.select_one(f'body > table.type2 > tbody > tr:nth-child({j}) > td:nth-child(5) > span').text.replace(',','')
					low_price = soup.select_one(f'body > table.type2 > tbody > tr:nth-child({j}) > td:nth-child(6) > span').text.replace(',','')
					sql = (id[0], date, closing_price, market_price, high_price, low_price)
					self.cur.execute('INSERT IGNORE INTO Stock_Price (stock_id , date, closing_price, market_price, high_price, low_price) VALUES (%s ,%s ,%s ,%s ,%s ,%s)', sql)
					self.conn.commit()
					break	
				else:
					continue				
				if date == '2022.03.31':
					break
		self.conn.close()
		self.driver.quit()