Programmiererforum

import time
import datetime as dt
import pandas as pd
from bs4 import BeautifulSoup
from dateutil.relativedelta import relativedelta
import requests
import re
import os

path = os.getcwd()

def ZCCWord(Date,country):

# Site URL
url="http://www.worldgovernmentbonds.com/country/"+country

html_content = requests.get(url).text
soup = BeautifulSoup(html_content, "lxml")
#gdp = soup.find_all("table", attrs={"class": "w3-table w3-white table-padding-custom w3 small font-family-arial table-valign-middle"})
gdp = soup.find_all("table") # , attrs={"class": "w3-table money pd44 -f15"})
table1 = gdp[0]
body = table1.find_all("tr")
body_rows = body[1:]
all_rows = [] # will be a list for list for all rows
for row_num in range(len(body_rows)): # A row at a time
row = [] # this will old entries for one row
for row_item in body_rows[row_num].find_all("td"): #loop through all row entries
aa = re.sub("(\xa0)|(\n)|,","",row_item.text)
#append aa to row - note one row entry is being appended
row.append(aa)
# append one row to all_rows
all_rows.append(row)

AAA           = pd.DataFrame(all_rows)
ZCC           = pd.DataFrame()
ZCC           = AAA[1].str.extract('([^a-zA-Z]+)([a-zA-Z]+)', expand=True).dropna().reset_index(drop=True)
ZCC.columns   = ['TENOR', 'PERIOD']
ZCC['TENOR'] = ZCC['TENOR'].str.strip().str.isdigit()  # Remove leading/trailing spaces
#ZCC = ZCC[ZCC['TENOR'].str.isdigit()]
ZCC['TENOR']  = ZCC['TENOR'].astype(int)
ZCC['RATES']  = AAA[2].str.extract(r'([0-9.]+)', expand=True).dropna().reset_index(drop=True).astype(float)
ZCC['RATES']  = ZCC['RATES']/100

row2      = []
for i in range(len(ZCC)):
if ZCC['PERIOD'][i]=='month' or  ZCC['PERIOD'][i]=='months':
b  = ZCC['TENOR'][i]
bb = Date + relativedelta(months = b)
row2.append(bb)
else:
b  = ZCC['TENOR'][i]
bb = Date + relativedelta(years = b)
row2.append(bb)

ZCC['DATES'] = pd.DataFrame(row2)
ZCC = ZCC.reindex(['TENOR','PERIOD','DATES','RATES'], axis=1)
return ZCC

LitsCountries   =  ['spain','portugal','latvia','ireland','united-kingdom',
'germany', 'france','italy','sweden','finland','greece',
'poland','romania','hungary','netherlands']

todays_date     = path+'\\WorldYields' +str(dt.datetime.now().strftime("%Y-%m-%d-%H-%M") )+ '.xlsx'
writer          = pd.ExcelWriter(todays_date, engine='xlsxwriter',engine_kwargs={'options':{'strings_to_urls': False}})
dictYield       = {}

for i in range(len(LitsCountries)):
country         = LitsCountries[i]
Date            = pd.to_datetime('today').date()
country         = LitsCountries[i]
ZCC             = ZCCWord(Date,country)
dictYield[i]    = ZCC
ZCC.to_excel(writer, sheet_name=country)

writer.close()
time.sleep(60) # wait one minute

Ich versuche, Renditetabellen für mehrere Länder und verschiedene Laufzeiten von einer Website zu extrahieren.
Bisher erhalte ich nur leere Tabellen:
[img]https://i.sstatic.net/md6bhS2D.png[/img]

während es eher so aussehen sollte:
[img]https://i.sstatic.net/Tp2yhYJj.png[/img]
Bisher habe ich Folgendes getan:
[code]import time
import datetime as dt
import pandas as pd
from bs4 import BeautifulSoup
from dateutil.relativedelta import relativedelta
import requests
import re
import os

path = os.getcwd()

def ZCCWord(Date,country):

# Site URL
url="http://www.worldgovernmentbonds.com/country/"+country

html_content = requests.get(url).text
soup = BeautifulSoup(html_content, "lxml")
#gdp = soup.find_all("table", attrs={"class": "w3-table w3-white table-padding-custom w3 small font-family-arial table-valign-middle"})
gdp = soup.find_all("table") # , attrs={"class": "w3-table money pd44 -f15"})
table1 = gdp[0]
body = table1.find_all("tr")
body_rows = body[1:]
all_rows = [] # will be a list for list for all rows
for row_num in range(len(body_rows)): # A row at a time
row = [] # this will old entries for one row
for row_item in body_rows[row_num].find_all("td"): #loop through all row entries
aa = re.sub("(\xa0)|(\n)|,","",row_item.text)
#append aa to row - note one row entry is being appended
row.append(aa)
# append one row to all_rows
all_rows.append(row)

AAA = pd.DataFrame(all_rows)
ZCC = pd.DataFrame()
ZCC = AAA[1].str.extract('([^a-zA-Z]+)([a-zA-Z]+)', expand=True).dropna().reset_index(drop=True)
ZCC.columns = ['TENOR', 'PERIOD']
ZCC['TENOR'] = ZCC['TENOR'].str.strip().str.isdigit() # Remove leading/trailing spaces
#ZCC = ZCC[ZCC['TENOR'].str.isdigit()]
ZCC['TENOR'] = ZCC['TENOR'].astype(int)
ZCC['RATES'] = AAA[2].str.extract(r'([0-9.]+)', expand=True).dropna().reset_index(drop=True).astype(float)
ZCC['RATES'] = ZCC['RATES']/100

row2 = []
for i in range(len(ZCC)):
if ZCC['PERIOD'][i]=='month' or ZCC['PERIOD'][i]=='months':
b = ZCC['TENOR'][i]
bb = Date + relativedelta(months = b)
row2.append(bb)
else:
b = ZCC['TENOR'][i]
bb = Date + relativedelta(years = b)
row2.append(bb)

ZCC['DATES'] = pd.DataFrame(row2)
ZCC = ZCC.reindex(['TENOR','PERIOD','DATES','RATES'], axis=1)
return ZCC

LitsCountries = ['spain','portugal','latvia','ireland','united-kingdom',
'germany', 'france','italy','sweden','finland','greece',
'poland','romania','hungary','netherlands']

todays_date = path+'\\WorldYields' +str(dt.datetime.now().strftime("%Y-%m-%d-%H-%M") )+ '.xlsx'
writer = pd.ExcelWriter(todays_date, engine='xlsxwriter',engine_kwargs={'options':{'strings_to_urls': False}})
dictYield = {}

for i in range(len(LitsCountries)):
country = LitsCountries[i]
Date = pd.to_datetime('today').date()
country = LitsCountries[i]
ZCC = ZCCWord(Date,country)
dictYield[i] = ZCC
ZCC.to_excel(writer, sheet_name=country)

writer.close()
time.sleep(60) # wait one minute
[/code]
Ich wäre auch mit anderen Websites, Lösungen oder Methoden zufrieden, die ähnliche Ergebnisse liefern.
Irgendeine Idee?
Vielen Dank im Voraus!