使用 Python 比較基金月份持股

最近看了介紹股票籌碼相關的節目，介紹可以透過基金持股來判斷，
基金的月報表會在每個月的第10個工作天公布上個前10大持股，
所以用 Python 抓了基金的網站來做最新兩個月的持股來做比較

使用
Python3
mechanicalsoup

	import numpy
	import mechanicalsoup
	import pandas as pd
	all_df = pd.DataFrame([], columns = ['Fund', 'Date', 'StockNo', 'StockName', 'Money'])
	all_df.set_index(['Date', 'Fund'], inplace = True)
	browser = mechanicalsoup.StatefulBrowser(
	soup_config={'features': 'lxml', 'from_encoding':'utf-8'},
	raise_on_404=True,
	user_agent='Mozilla/5.0 (compatible; MSIE 8.0; Windows NT 6.3; Win64; x64)'
	)
	req = browser.open("https://www.sitca.org.tw/ROC/Industry/IN2629.aspx")
	def dl_month ():
	soup = browser.get_current_page()
	options = soup.find("select", {"name": "ctl00$ContentPlaceHolder1$ddlQ_YM"}).find_all ('option')
	return [options[-2]["value"], options[-1]["value"]]
	def dl_page (month, calss_type):
	form = browser.select_form('#aspnetForm')
	form.form.find("select", {"name": "ctl00$ContentPlaceHolder1$ddlQ_Comid"})["disabled"] = "disabled"
	del form.form.find("select", {"name": "ctl00$ContentPlaceHolder1$ddlQ_Class"})["disabled"]
	form.form.find("select", {"name": "ctl00$ContentPlaceHolder1$ddlQ_Comid1"})["disabled"] = "disabled"
	form.form.find("select", {"name": "ctl00$ContentPlaceHolder1$ddlQ_Class1"})["disabled"] = "disabled"
	browser["ctl00$ContentPlaceHolder1$ddlQ_YM"] = month
	browser["ctl00$ContentPlaceHolder1$rdo1"] = "rbClass"
	browser["ctl00$ContentPlaceHolder1$ddlQ_Class"] = calss_type
	form.choose_submit("ctl00$ContentPlaceHolder1$BtnQuery")
	response = browser.submit_selected()
	#browser.launch_browser()
	soup = browser.get_current_page()
	tag = soup.select('table')[3].find_all ('td')
	tag_len = len (tag)
	df = pd.DataFrame([], columns = ['Fund', 'StockNo', 'StockName', 'Money'])
	fund_name = ""
	idx = 10
	while idx < tag_len:
	if tag[idx].text == "合計":
	fund_name = ""
	idx += 2
	continue
	if fund_name == "":
	fund_name = tag[idx].text.strip()
	idx += 1
	continue
	stock_type = tag[idx+1].text.strip()
	if stock_type != "國內上市" and stock_type != "國內上櫃":
	idx = idx + 9
	continue
	df = df.append ({
	'Fund': fund_name,
	"StockNo" : tag[idx+2].text,
	"StockName" : tag[idx+3].text,
	"Money" : numpy.int64 (tag[idx+4].text.replace(',', ''))},
	ignore_index=True)
	idx = idx + 9
	df['Date'] = month
	df.set_index(['Date', 'Fund'], inplace = True)
	return df
	months = dl_month ()
	for m in months:
	dl_page (m, 'AA1') #第一次會失效？？
	df1 = dl_page (m, 'AA1')
	df2 = dl_page (m, 'AB1')
	df3 = dl_page (m, 'AI1')
	df4 = dl_page (m, 'AH1')
	all_df = pd.concat([all_df, df1, df2, df3, df4], sort=True)
	#月別基金變動
	print ("月別基金變動")
	all_df.index.unique(level='Fund')
	index_fund = all_df.index.unique(level='Fund')
	for fund in index_fund:
	df = all_df[all_df.index.get_level_values('Fund').isin([fund])]
	prev_m = df[df.index.get_level_values('Date').isin([months[0]])]
	last_m = df[df.index.get_level_values('Date').isin([months[1]])]
	prev_stock = prev_m['StockNo'].tolist ()
	last_stock = last_m['StockNo'].tolist ()
	#保留持股
	intersection =list(set(prev_stock).intersection(last_stock))
	print (fund + " 移出前十持股")
	for s in prev_stock:
	if s not in intersection:
	stockname = df[df['StockNo'] == str(s)].StockName[0]
	print ("["+str(s)+"] "+ stockname)
	print (fund + " 新增前十持股")
	for s in last_stock:
	if s not in intersection:
	stockname = df[df['StockNo'] == str(s)].StockName[0]
	print ("["+str(s)+"] "+ stockname)
	print (fund + " 保留前十持股")
	for s in intersection:
	stockname = df[df['StockNo'] == str(s)].StockName[0]
	print ("["+str(s)+"] "+ stockname)
	#個股增減變動
	print ("個股增減變動")
	index_stock = sorted (list(set(all_df['StockNo'].tolist ())))
	prev_m = all_df[all_df.index.get_level_values('Date').isin([months[0]])]
	last_m = all_df[all_df.index.get_level_values('Date').isin([months[1]])]
	for s in index_stock:
	stockname = all_df[all_df['StockNo'] == str(s)].StockName[0]
	prev_data = prev_m[prev_m['StockNo'] == str(s)]
	prev_money = 0
	if not prev_data.empty:
	prev_money = prev_data.Money.sum()
	last_data = last_m[last_m['StockNo'] == str(s)]
	last_money = 0
	if not last_data.empty:
	last_money = last_data.Money.sum()
	if prev_money <= 0:
	print ("["+str(s)+"] "+ stockname + " 新增十大持股")
	elif last_money <= 0:
	print ("["+str(s)+"] "+ stockname + " 移出十大持股")
	elif last_money > prev_money:
	print ("["+str(s)+"] "+ stockname + " 十大持股增加")
	else:
	print ("["+str(s)+"] "+ stockname + " 十大持股減少")

view raw 基金月份比較 hosted with ❤ by GitHub